Generalized mixture operators using weighting functions: a comparative study with WA and OWA

Size: px
Start display at page:

Download "Generalized mixture operators using weighting functions: a comparative study with WA and OWA"

Transcription

1 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) Geeralzed mxture operators usg weghtg fuctos: a comparatve study wth WA ad OWA Rta Almeda Rbero *, Rcardo Alberto Marques Perera ** * Uversdade Lusíada, Departameto Cêcas Empresaras, Rua da Juquera 192, Lsboa, Portugal, rar@uova.pt ** Uverstà d Treto, Dpartmeto d Iformatca e Stud Azedal, Va Iama 5, TN Treto, Itala, mp@cs.ut.t Abstract. I the cotext of multple attrbute decso makg, we preset a aggregato scheme based o geeralzed mxture operators usg weghtg fuctos ad we compare t wth two stadard aggregato methods: weghted averagg (WA) ad ordered weghted averagg (OWA). Specfcally, we cosder lear ad quadratc weght geeratg fuctos that pealze bad attrbute performaces ad reward good attrbute performaces. A llustratve example, borrowed from the lterature, s used to perform the operators comparso. We beleve that ths comparatve study wll hghlght the potetal ad flexblty of geeralzed mxture operators usg weghtg fuctos that deped o attrbute performaces. Keywords: fuzzy sets, weghted aggregato, lear ad quadratc weght geeratg fuctos, geeralzed mxture operators, multple attrbute decso makg. 1. Itroducto I multple attrbute decso makg problems for a gve umber of alteratves or courses of acto, oe has to detfy a set of relevat attrbutes characterzg the alteratves, elct ther relatve mportace ad express these some form of weght factors. The usual uderlyg assumpto s that attrbutes are depedet,.e. the cotrbuto of ay dvdual attrbute to the total ratg of a alteratve s depedet of the other attrbute values. I ths paper, we assume that a umercal characterzato of all relevat attrbutes s possble ad les wth the famlar ut scale [0,1]. Moreover, as most of the research o weghted aggregato, from the classcal weghted averagg (WA) ad the ordered weghted averagg (OWA) (Yager 1988) (Yager 1992; Yager 1993) to the more recet Choquet tegrato (over a fte doma) 1

2 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) (Grabsch 1995; Grabsch 1996), we assume commesurablty of the varous attrbutes wth the ut scale. Ths meas that the way attrbute satsfacto values are umercally expressed the ut terval s desged so that commesurablty holds to a acceptable degree. We also assume that the relatve mportace of attrbutes ca tself be expressed umercally wth the ut terval. I solvg multple attrbute decso makg problems may types of soluto methods have bee developed (see for stace (Yoo ad Hwag 1995)): scorg ad outrakg methods, trade-off schemes, dstace based methods, value ad utlty fuctos, teractve methods. The frst exteso of multple attrbute decso models to the fuzzy set cotext was proposed by Bellma ad Zadeh (Bellma ad Zadeh 1970). Sce the most of the techques to solve multple attrbute problems have bee exteded to the fuzzy doma (for good overvews see (Che ad Hwag 1992) (Zmmerma 1987) (Kckert 1978) (Rbero 1996)). Scorg techques are wdely used, partcularly the weghted aggregato operators based o multple attrbute value theory (Keeey ad Raffa 1976). The classcal weghted aggregato s usually kow the lterature by weghted averagg (WA) or smple addtve weghtg method (Yoo ad Hwag 1995). Aother mportat aggregato operator, wth the class of weghted aggregato operators, s the ordered weghted averagg OWA (Yager 1988). I ths paper we wll focus o two famles of geeralzed mxture operators (Marques Perera ad Pas 1999; Marques Perera 2000) usg weghtg fuctos that pealze bad attrbute performaces ad reward good attrbute performaces: oe famly s assocated wth lear weght geeratg fuctos ad the other famly s assocated wth quadratc weght geeratg fuctos. Ths aggregato method exteds weghted averagg, whch s a partcular case of geeralzed mxture operator. The questo of defg weghts (.e. weghtg fuctos) that deped o attrbute satsfacto values as (Rbero ad Marques Perera 2001) s at the kerel of ths work ad we beleve t ca be relevat some decso problems. For stace, suppose that the parets of a 5 th grade studet wsh to select a school for ther so. There are two caddate schools ad the parets wat to select the best oe accordg to the followg two crtera. The most mportat crtero for school selecto s dstace ; small dstace meas that school s wth walkg dstace from home, whereas large 2

3 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) dstace meas that the school ca be reached oly by car. The secod (less mportat) crtero s the teachg qualty of the school. The decsoal scheme of the parets s as follows: f both schools are close from home (wth walkg dstace) tha domace w.r.t. dstace s much stroger tha domace w.r.t. qualty, sce the dfferece dstace s expereced walkg. O the other had, f both schools are far from home (ca be reached oly by car) tha domace w.r.t. dstace s comparable wth domace w.r.t. qualty, sce the dfferece dstace s expereced by drvg. We model ths decsoal scheme by cosderg lear weght geeratg fuctos of the followg types: very mportat (wth terval [0.6, 1]) for dstace ad mportat (terval [0.4, 0.8]) for qualty. For the weghted average (WA) we cosder the mddle pot of the weght rages, respectvely 0.8 ad 0.6. Now, Table 1 we preset two possble selecto cases (Case I ad Case II) wth attrbute satsfacto values ad aggregated values for each of the schools uder cosderato, usg the two lear weght geeratg fuctos dcated before. CASE I School 1 School 2 Dstac e Qualt y RESULT S CASE School School II 1 2 Dstac e Qualt y RESULT S Table 1. Example of school selecto wth attrbutes dstace ad qualty To clarfy the use of lear weghtg fuctos computg the aggregated values, cosder that are the weght geeratg fuctos for attrbutes dstace ad qualty, respectvely. The vector x = ( x D, f ( x ) = (1 0.6) x the aggregated value of the school s gve by D x Q ) D D ad f ( x ) = ( ) x represets the attrbute satsfacto values of oe school ad Q Q Q ( x ) = = D, Q w ( x x W ) = j = D Q wth weghtg fuctos w (x) f ( x ) /, f ( x ). (1) For stace, Case I the aggregated value of school 1 s computed as follows, j j 3

4 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) Case1 school 1 = 0.4 * (0.6 + (1 0.6) * 0.4) * (0.4 + ( ) * 0.8) = (0.6 + (1 0.6) * 0.4) + (0.4 + ( ) * 0.8 Observg Table 1 we ca see that for Case I the selected school s 2, whle Case II the selected school s 1. Ths smple example shows clearly the depedecy of weghts o the satsfacto values of attrbutes, rewardg hgher values of attrbute satsfacto ad pealzg lower values. I both Cases I ad II, there s a domace of 0.2 oe attrbute agast a detcal but reverse domace of 0.2 the other attrbute. I Case II the domace the attrbute dstace ws, whch s just what oe would expect gve that dstace s the most mportat attrbute. I Case I, o the other had, t s the domace the attrbute qualty that ws, because ths case the dstace satsfacto values are low ad therefore the effect of the domace that attrbute s pealzed. Cosderg the chage from Case I to Case II we see that, sce dstace s the most mportat attrbute, whe ts attrbute satsfacto values crease from low to hgh the school choce chages. I other words, attrbutes wth bad performaces are pealzed ad attrbutes wth good performaces are rewarded, wth the correspodg effect o attrbute domace. O the other had, f we apply ether the classcal weghted averagg (WA) or the ordered weghted averagg (OWA) to the school example, the selected school s always 2. Ths s clear the WA case; as far as the OWA case s cocered, ths s because the weght assocated wth attrbute dstace s the same for both schools ad does ot chage from Case I to Case II (otce that the attrbute satsfacto value for dstace s always less or equal tha that of the attrbute qualty ). I summary, the ma focus of ths work s to llustrate the potetal ad flexblty of the aggregato wth geeralzed mxture operators usg weghtg fuctos, by comparso wth two other aggregato methods, the weghted averagg WA ad the OWA (Yager 1988). The paper s orgazed as follows. After ths troductory secto, secto 2 provdes a bref troducto to the three aggregato methods that wll be compared. Secto 3 presets the ma cocepts volved weghted aggregato ad secto 4 descrbes the lear ad quadratc weght geeratg fuctos that wll be used wth the geeralzed mxture operators. Secto 5 compares ad dscusses the three approaches 4

5 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) usg a llustratve example, borrowed from the lterature. Fally, secto 6 presets the coclusos of ths comparatve study. 2. Weghts ad weghtg fuctos The specfcato of a preferece rakg over a set of attrbutes reflects the relatve mportace that each attrbute has the composte. Procedures for elctg ad determg relatve mportace (weghts) have bee the focus of extesve research ad dscusso (Al-Kloub, Al-Shemmer et al. 1997), (Edwards ad Newma 1982) (Hobbs 1980), (Rbero 1996), (Stllwell, Seaver et al. 1981), (Weber ad Borcherdg 1993), (Yoo 1989). However, a mportat problem that has oly bee partally addressed by the research o elctg ad determg weghts s the fact that weghts should sometmes deped o the correspodg attrbute satsfacto values. From a decso maker perspectve, whe cosderg oe attrbute wth a hgh relatve mportace, whch has a bad performace (low satsfacto value), the decsoal weght of ths attrbute should be pealzed, order to reder the gve attrbute less sgfcat the overall evaluato of the alteratves. As a result, whe cosderg two alteratves, the domace effect oe mportat attrbute becomes less sgfcat whe the attrbute satsfacto values are low. I the same sprt, the decsoal weght of a attrbute wth low relatve mportace (weght) that has hgher satsfacto values should be rewarded, thereby rederg more sgfcat the domace effect that attrbute. Summarzg, some cases t may be approprate to have weghts depedg o the attrbute satsfacto performace values. The troducto of weghtg fuctos depedg cotuously o attrbute satsfacto values produces aggregato operators wth complex umercal behavour. The mootocty of the aggregato operator s a crucal ssue (Marques Perera ad Pas 1999; Marques Perera 2000) whch volves costrats o the dervatves of the weghted aggregato operator wth respect to the varous attrbute satsfacto values. Naturally, the questo of mootocty s also relevat the classcal framework of umercal weghts, partcularly whe the weghted aggregato operators have complex x depedeces. I ths case, moreover, there s also the questo of sestvty, volvg costrats o the dervatves of the aggregato operator wth respect to the varous weght values. Roughly speakg, whle mootocty requres that the operator 5

6 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) creases whe ay of the attrbute satsfacto values creases, sestvty requres that the relatve cotrbuto of the -th attrbute to the value of the operator creases whe the correspodg weght creases. A terestg study of mootocty ad sestvty the classcal framework of umercal weghts has bee doe (Kaymac ad va Nauta-Lemke 1998). A alteratve approach to modulate the relatve mportace of attrbutes depedg o attrbute satsfacto values was proposed by Rbero (Rbero 2000). The author troduced a lear terpolato fucto, called WIS (Weghtg Importace ad Satsfacto), whch combes lgustc mportace weghts of attrbutes wth attrbute satsfacto values. The lgustc relatve mportace values are represeted by fuzzy sets (Zadeh 1987) o a cotuous terval support, as for stace very_mportat=[0.6, 1], ad attrbute satsfacto values are represeted the terval [0.1,1]. Recetly, wth the geeralzed mxture approach, two weghtg fuctos depedg o attrbute satsfacto values were proposed (Rbero ad Marques- Perera 2001). Oe s obtaed from a lear weght geeratg fucto ad the other s obtaed from a sgmod weght geeratg fucto. I both cases the relatve mportace (weghts) of the attrbutes s provded lgustc format, as for example {very mportat, mportat, ot mportat}, ad each lgustc weght has a lower ad upper lmt gve by the decso maker. The lower ad upper lmts are the tegrated the weght geeratg fuctos to defe ther rage ad slope. I ths paper we develop further the geeralzed mxture approach by cosderg a quadratc exteso of the lear weght geeratg fucto. The ext two sectos descrbe the ma characterstcs of the two weght geeratg fuctos (lear ad quadratc) cosdered. 3. Weghted aggregato operators: WA, OWA, ad geeralzed mxtures I ths secto we brefly recall the defto ad ma propertes of three types of weghted aggregato operators: weghted averagg (WA) operators, ordered weghted averagg (OWA) operators, ad geeralzed mxture operators. The latter exted the classcal weghted averagg operators (whch rema a specal case) the sese that the classcal costat weghts become weghtg fuctos defed o the 6

7 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) aggregato doma. I other words, the weghts are o loger costat but deped o the values of the aggregato varables. The presetato of geeralzed mxtures follows the orgal oes (Marques Perera ad Pas 1999; Marques Perera 2000), where mxture operators ad geeralzed mxture operators have bee troduced. Roughly speakg, mxture operators are cotuous ad compesatve aggregato operators. I other words, they are mea operators wthout the mootocty requremet. Geeralzed mxture operators costtute a partcular class of mxture operators, the same way as geeralzed mea operators (also called quas-arthmetc meas) costtute a partcular class of mea operators. There s fact a strog ad terestg relato betwee the two geeralzed classes, as has bee poted out (Marques Perera ad Pas 1999) Weghted averagg (WA) operators Ths aggregato model, whose relevace stems from the axomatc bass of multple attrbute value theory (Keeey ad Raffa 1976) together wth the cetral role of weght determato multple value aalyss (Weber ad Borcherdg 1993), s also kow the lterature by smple addtve weghtg method (Yoo ad Hwag 1995). Usually, the WA model assumes that weghts are proportoal to the relatve value of a ut chage each attrbute (Hobbs 1980). Cosder varables the ut terval [0,1], wth = 1,...,, ad a equal umber of weghts 0 satsfyg the ormalzato codto = 1. The w vectors cotag the varables ad the weghts are deoted x = [ x ] ad w = [ w ], respectvely. where x j = w 1 j The classcal weghted averagg operator s defed as WA (x) = =1 w x, x = [ x ] ecodes the attrbute satsfacto values of a alteratve. The operator WA s clearly cotuous the aggregato doma [ 0,1]. The weghted averagg operator WA (x) s also compesatve ad mootoc, meag that m( x) WA ( x) max(x) ad x y WA( x) WA( y), 7

8 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) respectvely. The relato x y meas that x y for = 1,...,. Moreover, the WA(x) operator s lear ad the weghts are costat the aggregato doma Ordered weghted averagg (OWA) operators The ordered weghted averagg (OWA) operators have bee troduced (Yager 1988; Yager 1992; Yager 1993). Although the orgal OWA scheme dd ot clude quatfers, ths paper we use quatfed OWA operators (Yager 1996) for our comparatve study. The llustratve example dscussed the fal secto was borrowed from the work (Yager ad Kelma 1999). The ordered weghted averagg operator s defed as OWA ( x) = = 1wx( ), where x ( ) s the th largest elemet of the vector operator WA s cotuous the aggregato doma [ 0,1]. x. Despte the reorderg mechasm, the It s useful to rewrte the weghted averagg operator as OWA (x) = =1 w[ ] x, where w [] s the weght assocated wth the th elemet of the vector x, referrg to the tradtoal OWA operator defto gve above. I ths formulato of the reorderg mechasm t s evdet that the weght assocated wth a elemet of the vector x chages depedg o the order statstcs of that elemet. The ordered weghted averagg operator OWA (x) s also compesatve ad mootoc, that s m( x) OWA( x) max( x) ad x y OWA( x) OWA( y), respectvely. However, the OWA (x) operator s o lear ad the weghts are oly pecewse costat the aggregato doma: they are costat wth comootocty domas but chage dscotuously from oe comootocty doma to aother. We recall that comootocty domas are coes dmesoal aggregato space wth whch all vectors x have the same orderg sgature. For stace, for = 3 the two vectors x = (0.2,0. 7,0.5) ad x'= (0.3,0.9,0.8) have the same orderg sgature ((1),(2),(3))=(2,3,1) ad thus belog to the same comootocty doma. Istead, the vector x"= (0.6,0.1,0.2) wth orderg sgature ((1),(2),(3))=(1,3,2) belogs to a dfferet comootocty doma. 8

9 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) w (x) As we shall see below, geeralzed mxture operators have weghtg fuctos that deped cotuously o the aggregato varables x. Ths meas cosderg a large class of weghtg fuctos cotag, partcular, arbtrarly good approxmatos of the pecewse costat weghtg fuctos used the OWA operator. Apart from cotuty, t s coveet to assume also some degree of dfferetablty, so that the powerful techques of dfferetal calculus ca be appled. These techques are fact crucal characterzg the geeral mootocty costrats uder whch weghtg fuctos effectvely produce mootoc weghted aggregato operators (Marques Perera ad Pas 1999; Marques Perera 2000). We ote that a partcular class of geeralzed mxture operators had already bee cosdered (Yager ad Flev 1994) but volatos of mootocty had led the authors to abado ths proposal. Now we eed to troduce a lk betwee the OWA operator defto ad ts relatoshp wth lgustc quatfers Q, such as Most. Gve a lgustc quatfer Q, meag for stace that Q of the attrbutes should be satsfed by a acceptable soluto, the weghts ca be determed as (Yager 1996), 1 w ( x ) = Q( u ) Q( u ) (2) j= 1 ( j) j= 1 ( j) j =1 u j = 1 where the u are the ormalzed mportaces of attrbutes, wth 2 quatfer Most s gve by ( r ) r for r 0, 1. Q = [ ], ad the 3.3. Geeralzed mxture operators Let us cosder ow fuctos f (x) defed the ut terval x [0,1], wth = 1,...,. We assume that these fuctos are of class C,.e. each fucto 1 f (x) [ s cotuous ad has a cotuous dervatve f '( x) the ut terval x 0,1], for = 1,...,. We also assume that the fuctos (x) f are postve the ut terval,.e. f ( x) > 0 for x [0,1] ad = 1,...,, ad that ther dervatves '( x) are oegatve the ut terval,.e. '( x) 0 for x [0,1] ad = 1,...,. f f 9

10 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) (x) f 2000), The geeralzed mxture operator W (x) o [ 0,1] geerated by the fuctos, wth = 1,...,, s defed as (Marques Perera ad Pas 1999; Marques Perera W f ( x ) x = 1 (x) = (3) j= 1 f j ( x j ) 1 It s straghtforward to show that W (x) s also C the aggregato doma [ 0,1],.e. t s cotuous ad has cotuous partal dervatves [ 0,1]. Moreover, the geeralzed mxture operator W (x) ca be wrtte as a weghted average of the varables x, wth weghtg fuctos w (x) stead of the classcal w costat weghts, W ( x ) = = 1w ( x) x wth weghtg fuctos w (x) f ( x ) = j = 1 f j ( x j ) (4) Due to the propertes of the geeratg fuctos (x), for = 1,...,, the weghtg f fuctos are strctly postve, w ( x) > 0, ad satsfy the ormalzato codto j=1 w j ( x) = 1. Geeralzed mxture operators are compesatve operators the usual sese, but they are ot ecessarly mootoc. Nevertheless, terestg suffcet codtos for mootocty ca be derved, as show below. Moreover, geeralzed mxture operators are o lear (eve wth comootocty domas) ad ther weghtg fuctos vary cotuously ad smoothly the aggregato doma. I (Marques Perera ad Pas 1999; Marques Perera 2000), the authors have derved a geeral suffcet codto for the strct mootocty of the geeralzed mxture operator W (x), where by strct mootocty we mea ( x) > 0 for = 1,...,. I the ut terval, for postve geeratg fuctos wth o-egatve W 10

11 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) dervatves, the suffcet codto takes the smple form ad = 1,..., f '( x) f ( x), for. I the paper, we call t smply the mootocty codto. x [0,1] From a tutve pot of vew, the mootocty codto establshes a upper boud for the varablty of the weght geeratg fucto, f '( x) / f ( x) 1. I tur, ths costrat lmts the degree to whch the weghtg fuctos (x) w deped o the aggregato varables. I other words, the weght depedece o the aggregato varables must be some sese small ( the classcal weghted averagg case t was ull) so that mootocty s ot volated. Notce that the codtos 0 f '( x) f ( x) > 0 hold depedetly for each = 1,...,. We shall make use of ths fact what follows by applyg those codtos to a geeral lear or quadratc geeratg fucto f (x), wth oly o-classcal parameters regardg the weght depedece o the attrbute satsfacto degrees. Later we cosder the assocated effectve geeratg fucto F( x) = α f ( x) / f ( 1), whose value for ut attrbute satsfacto F(1) = α s gve by the classcal mportace parameter α. I secto 4 we descrbe two dfferet ways whch geeralzed mxture operators ca model multple attrbute aggregato wth weghtg fuctos: oe volves lear weght geeratg fuctos ad the other volves quadratc weght geeratg fuctos. 4. Lear ad quadratc weght geeratg fuctos I ths secto we descrbe weght geeratg fuctos of the lear ad quadratc types. A prelmary study of the lear case has appeared (Rbero ad Marques Perera 2001). I geeral, as dscussed the prevous secto we assume that the weght 1 geeratg fuctos, f (x) for x [0,1] ad = 1,...,, are of class C ad dvdually satsfy three geeral codtos: (I) f ( x) > 0, (II) '( x) 0, ad (III) f '( x) f ( x). The latter s called the mootocty codto ad plays a cetral role the costructo. f 11

12 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) We recall that the geeralzed mxture operator costructed from the geeratg fuctos s defed as W ( x) = = 1w ( x) x, where the weghtg fuctos j = w (x) = f ( x ) / 1 f ( x j j ) satsfy the usual ormalzato codto j=1 w j ( x) = 1 for all x [0,1] Lear weght geeratg fuctos Cosder the lear geeratg fucto l ( x) = 1+ βx, wth doma x [0,1] ad parametrc rage 0 β = 0. β 1. The classcal case, wth costat weghts, correspods to It s straghtforward to check that the above codtos (I-III) hold the parametrc rage 0 β 1. The frst two codtos, l ( x) > 0 ad l' ( x) 0, are clearly verfed. The mootocty codto ca be wrtte as β ( 1 x) 1. Accordgly, sce ( 1 x) [0,1] for x [0,1], the mootocty codto s also verfed. l' ( x) l(x) The effectve geeratg fucto L (x) assocated wth l (x) s gve by l( x) 1+ βx L ( x) = α = α (5) l(1) 1+ β wth parameters 0 < α 1 ad 0 β 1. The mmum ad maxmum values L (0) ad L(1) of the effectve geeratg fucto, wth 0 < L(0) L(1) 1, are gve by L( 0) = α /( 1+ β ) ad L(1) = α, respectvely. Parameter 0 < α 1 cotrols the maxmum value L(1) = α of the effectve geeratg fucto, whe the attrbute satsfacto value s oe. I tur, parameter 0 β 1 cotrols the rato L( 1) / L(0) = 1+ β betwee the maxmum ad mmum values of the effectve geeratg fucto. I ths respect, the parametrc codto 0 β 1 ca be wrtte the form 1 L( 1) / L(0) 2, meag that the rato L( 1) / L(0) s at most 2 the lear case. 12

13 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) I the classcal weghted averagg case, the effectve geeratg fuctos have ull parameter β ( ths way they do ot deped o attrbute satsfacto) ad the dfferet costat weghts are obtaed by dfferet choces of the parameter α. Lear wth alpha=0.8 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0, ,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 beta=0 beta=0.5 beta=1 Fgure 1. Lear weght geeratg fuctos Fgure 1 above shows three effectve geeratg fuctos of the lear type, assocated wth α = 0.8 ad three dfferet choces of parameter β. Parameter α cotrols the maxmum value of the effectve geeratg fucto, whe the attrbute satsfacto value s 1, whle parameter β cotrols the rato betwee the maxmum ad mmum values of the geeratg fucto Quadratc weght geeratg fuctos I complete aalogy wth the lear case descrbed prevously, oe ca study geeratg fuctos depededg quadratcally o attrbute satsfacto values. Cosder thus the quadratc geeratg fucto q( x) = 1+ γ x 2, wth doma x [0,1] ad parametrc rage 0 γ 1. The classcal case, wth costat weghts, correspods to γ = 0. As before, t s straghtforward to check that codtos (I-III) hold the parametrc rage 0 γ 1. The frst two codtos, q( x) > 0 ad q' ( x) 0, are clearly verfed. I tur, the mootocty codto q '( x) q( x) ca be wrtte as 13

14 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) γ (2x x ) 1. Accordgly, sce (2x x ) [0,1] for x [0,1], the mootocty codto s also verfed. The effectve geeratg fucto Q (x) assocated wth q (x) s gve by q( x) Q(x) = α q(1) 1+ γ x = α 1 + γ 2 (6) wth parameters 0 < α 1 ad 0 γ 1. As the prevous case, the mmum ad maxmum values Q (0) ad Q (1) of the effectve geeratg fucto, wth 0 < Q(0) Q( 1) 1, are gve by Q( 0) = α /(1 + γ ) ad Q (1) = α, respectvely. Parameter 0 < α 1 cotrols the maxmum value Q(1) = α of the effectve geeratg fucto, whe the attrbute satsfacto value s oe. I tur, parameter 0 γ 1 cotrols the rato Q (1) / Q(0) = 1+ γ betwee the maxmum ad mmum values of the effectve geeratg fucto. I ths respect, the parametrc codto 0 γ 1 ca be wrtte the form 1 Q(1) / Q(0) 2, meag that the rato Q ( 1) / Q(0) s at most 2 also the quadratc case. As before, the classcal weghted averagg case the effectve geeratg fuctos have ull parameter γ ( ths way they do ot deped o attrbute satsfacto) ad the dfferet costat weghts are obtaed by dfferet choces of the parameter α. The two cases, lear ad quadratc, are equvalet most respects, partcularly the aalogous role of the parameters β,γ ad the detcal rage (from 1 to 2) of the rato betwee the maxmum ad mmum values of the effectve geeratg fuctos. The real dfferece betwee the lear ad quadratc cases les the fact that the lear depedecy o attrbute satsfacto values s homogeeous the doma x [0,1], whereas the quadratc depedecy s weak for low attrbute satsfacto values x [ 0, 1 2] ad strog for hgh attrbute satsfacto values x [ 1 2,1]. Ths homogeeous depedecy o attrbute satsfacto values has the effect of 14

15 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) emphaszg the o classcal behavour of the assocated geeralzed mxture operators, as wll be llustrated the example preseted the last secto of the paper. Regardg the mportat questo of the rato betwee the maxmum ad mmum values of the effectve geeratg fuctos, t s possble to sgfcatly exted the preset 1 to 2 rage by cosderg a more geeral famly of quadratc 2 geeratg fuctos, cotag also a lear term: q( x) = 1+ ( β γ ) x + γ x. Roughly speakg, ths exteded geeratg fucto the presece of parameter β ehaces the effect of parameter γ ad (wth approprate parameter tug) the rato Q( 1) / Q(0) ca reach up to 2.6, approxmately. The theoretcal dscusso of ths possblty, however, s beyod the scope of ths paper ad wll be preseted elsewhere. Lear & Quadratc wth alpha=0.8 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0, ,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Lbeta=0 Lbeta=0.5 Lbeta=1 Qgamma=0 Qgamma=0.5 Qgamma=1 Fgure 2. Lear ad quadratc weght geeratg fuctos Fgure 2 above shows three effectve geeratg fuctos of the quadratc type, assocated wth α = 0.8 lear geeratg fuctos wth ad three dfferet choces of parameter β. The correspodg β = γ are also depcted. As before, parameter α cotrols the maxmum value of the effectve geeratg fucto, whe the attrbute satsfacto value s 1, whle parameter γ cotrols the rato betwee the maxmum ad mmum values of the geeratg fucto. 15

16 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) Illustratve example Ths example s borrowed from Yager ad Kelma (Yager ad Kelma 1999) ad cossts the problem of selectg the best caddate for a computer developer job. There are three caddates ad ther evaluato s made o the bass of sx attrbutes. We dscuss ths decsoal problem wth the followg techques: Yager s quatfed OWA operator (Yager 1988); the classcal weghted averagg operator; ad the geeralzed mxture operators assocated wth our lear ad quadratc weght geeratg fuctos. We compare the results obtaed wth the four techques order to llustrate the flexblty of our weght geeratg fuctos the cotext of decsoal problems whch depedece o attrbute satsfacto s a relevat ssue. As metoed before, there are three caddates for the computer developer job ad the decso maker evaluates them wth respect to sx attrbutes. The sx attrbutes are the computer laguage kowledge of: Basc, Pascal, Lsp, Prolog, C++, ad Vsual Basc. Moreover, the decso maker thks that some laguages are more mportat tha others ad assgg dfferet mportace values to dfferet attrbutes wll represet ths. For what cocers the quatfed OWA method, the oly dfferece betwee our dscusso of the computer developer problem ad the orgal oe descrbed (Yager ad Kelma 1999) s that we wat to select the caddate wth the best kowledge of the most mportat laguages, stead of selectg the caddate wth the best kowledge of most laguages. For ths reaso, we use the lgustc quatfer Few stead of the lgustc quatfer Most orgally used (Yager ad Kelma 1999). Mathematcally, we represet the lgustc quatfer Few by the ut terval [ ] trasformato Q ( r ) = r for r 0, 1. Ths substtuto of the lgustc quatfer Most wth Few eables us to make a more meagful comparso betwee our geeralzed mxtures ad the quatfed OWA operators sce, ths paper, we oly study the case of creasg weght geeratg fuctos. Ths class of geeratg fuctos produces larger weghts for larger attrbute satsfacto values, whch s roughly speakg the average behavour of quatfed OWA operators based o the lgustc quatfer Few. If o the other had we were studyg decreasg weght geeratg fuctos we could have kept the lgustc quatfer Most, as the orgal verso of the decsoal problem. 16

17 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) I Table 2 we show the orgal data for the decsoal problem questo (see also (Yager ad Kelma 1999)). It cludes the satsfacto values of each attrbute, for the three caddates, as well as the mportace values of the sx attrbutes. Importace Attrbutes Basc Pascal Lsp Prolog C++ Vsual Basc Caddate A Caddate B Caddate C Table 2. Data for the computer developer problem The frst step to solve ths decsoal problem s to trasform the mportace values to a [0,1] scale, sce our effectve weght geeratg fuctos are o that scale. Hece, we dvde the mportace of each attrbute by the maxmum mportace value. These scaled mportace values wll be used throughout the dscusso. Table 3 shows the orgal mportace values ad the scaled oes. Attrbutes Basc Pascal Lsp Prolog C++ Vsual Basc Importace Scaled Importace Table 3. Scaled mportace values The scaled mportace values are used our lear ad quadratc weght geeratg fuctos to defe the parameter α, whch correspods to the upper lmt of the fuctos. Both parameters β,γ for the lear ad quadratc geeratg fuctos (respectvely) are set to 1, so as to have the maxmal slope both cases. Furthermore, lgustc terms we ca terpret the attrbute mportace values as {very_low, low, hgh, very_hgh} ad these terms correspod to the upper lmts α ={0.2, 0.4, 0.6, 1}, respectvely. I Fgure 3 we show the lear ad quadratc geeratg fuctos for the weghts whose mportace s hgh ad very_hgh (Pascal, C++, ad Vsual Basc). 17

18 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) Weghts Hgh ad Very-Hgh 0,9 1 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0, ,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Lear(0.6, 1) Lear(1, 1) Quadratc(0.6,1,1) Quadratc(1,1,1) Fgure 3. hgh: Lear(0.6,1), Quadratc(0.6,1); very_hgh: Lear(1,1), Quadratc(1,1). The remag weght geeratg fuctos used our dscusso have upper lmts α = 0.2 ad α =0.4, correspodg to very_low ad low mportace values. For what cocers the WA ad OWA operators used ths example, the costructo s as follows. The weghts the WA operator are gve by the ormalzed (ut sum) mportace values computed from Table 3,.e. the weght for Pascal s 0.2/3.6 ad so o. The same values of ormalzed mportace are also used the OWA operator costructo. I ths case they play the role of the u, for = 1,...,, whch eter the defto of the OWA weghts through the quatfer scheme descrbed at the ed of secto 3.2. We are ow the posto to preset the results usg the four methods: the classcal weghted averagg operator WA; Yager s quatfed OWA aggregato operator; ad the geeralzed mxture operators assocated wth our lear ad quadratc weght geeratg fuctos. Table 4 shows the aggregato results obtaed wth the four methods. Lear Quadratc WA OWA Caddate A Caddate B Caddate C Table 4. Aggregato results Observg Table 4 we ca see that all techques select caddate A as the best oe, ad that the weghted averagg operator WA fals to detect the secod best 18

19 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) caddate B. Both the OWA ad our lear ad quadratc weght geeratg fuctos select caddate B as the secod best ad caddate C as the worst caddate. I order to llustrate the flexblty of our weght geeratg fuctos, whch deped o the satsfacto values of the attrbutes, we made a smulato cosderg that caddate A had a lower attrbute satsfacto value for laguage C++. More specfcally, we wated to see what would happe f caddate A had a satsfacto value of 0.2 stead of 0.5 for a attrbute wth very_hgh mportace. The correspodg aggregato results are show table 5. Lear Quadratc WA OWA Caddate A Caddate B Caddate C Table 5. Aggregato results usg attrbute C++ wth 0.2 (stead of 0.5) for caddate A. Observg the results obtaed Table 5 we ca see that, ow, the best caddate s B for our lear ad quadratc weght geeratg fuctos, whereas the weghted averagg operator WA does ot dstgush betwee caddate B ad C, ad the quatfed OWA operator stll thks that caddate A s the best oe. However, f we observe the orgal attrbute satsfacto values (Table 2), wth the chage of 0.2 for attrbute C++ o caddate A, t s clear that caddate B performs much better (o average) attrbutes wth hgh ad very_hgh mportace, respectvely Pascal, C++, ad Vsual Basc: wth the orgal data (Table 2), the weghted average values of attrbute satsfacto restrcted to the three mportat attrbutes are A > B > C 0.615; o the other had, whe the attrbute satsfacto value of caddate A for C++ s chaged from 0.5 to 0.2, the weghted average values become B > C > A 0.515, wth A as the worst caddate ow. However, ether the WA or the OWA are able to take to accout ths fact. Ths smulato clearly shows that our weght geeratg fuctos ca pealze (or reward) alteratves that have lower (or hgher) satsfacto values for the attrbutes, partcularly whe the attrbute has a hgh or very_hgh mportace. Aother mportat aspect to study s the comparso betwee the actual weghtg fuctos of our geeralzed mxture operators ad the OWA weghts. Hece, we made 19

20 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) a smulato that llustrates the behavour of our weghtg fuctos ad the OWA weghts whe the satsfacto value of caddate A for attrbute C++ creases the terval [0.2, 0.5]. The umercal results are show Fgure 4. C++ weght comparso Lear Quadratc OWA WA Fgure 4. Comparso betwee our weghtg fuctos ad the OWA weghts Observg Fgure 4, otce that the OWA weght s costat for the terval cosdered whle our weghtg fuctos reward the crease the satsfacto value of the attrbute C++. The lear weghtg fucto has a faster crease the the quadratc oe, as could also be see Fgure 4. The choce betwee the lear ad quadratc weght geeratg fuctos should be left to the decso maker because t depeds o the problem, partcularly o how much the decso maker wshes to pealze or reward chages attrbute satsfacto. A fal smulato was made to study volatos of preferetal depedece (Keeey ad Raffa 1976) beyod those that occur the OWA aggregato scheme. Ths meas lookg for preferetal depedece volatos wth comootocty domas, whch would ecessarly mply that our geeralzed mxtures are able to represet preferece structures that ca ot be represeted by OWA operators. Cosder a vector x = x ] of attrbute satsfacto values, wth x [0,1]. As [ the OWA operator defto, we deote by () the dex value of the th largest elemet of the vector x. More precsely, we say that (( 1), (2),..., ( ) ) s a decreasg x ) orderg sgature assocated wth f x ( 1) x(2... x( ). I what follows the decreasg specfcato wll always be mplct ad s therefore omtted. The orderg 20

21 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) sgature s clearly ot uque whe the vector x has equal elemets. I geeral, we deote by π (x) the set of orderg sgatures assocated wth x. We say that two-attrbute satsfacto vectors x = [ x ] ad y = [ y ] [ 0,1] are comootoc f the assocated orderg sgatures π (x) ad π (y) cocde (as sets). Comootocty s thus a equvalece relato, whch parttos the aggregato doma [0,1] to mutually dsjot comootocty domas. Geometrcally, each comootocty doma s a coe from the org bouded by the lmts of the aggregato doma [ 0,1]. Now, cosder two vectors x = ] ad [ x 1,..., x 1, t, x + 1,..., x y = [ y y, 1,..., 1 t, y + 1,..., y ] [ 0,1] wth the same th elemet. We say that a aggregato operator W o [ 0,1] satsfes preferetal depedece wth respect to the th attrbute f, gve ay two reduced vectors x \ = [ x 1,..., x 1,x+ 1,..., x ] ad y\ = [ y 1,..., y 1, y+ 1,..., y ], we obta the same relato W ( x ) W( y ) or W( x ) W( y ) for all values of t [0,1]. We say that the aggregato operator W satsfes preferetal depedece f t satsfes preferetal depedece wth respect to all attrbutes. It s well kow that the weghted averagg WA operator satsfes preferetal depedece ad that the weghted averagg operator OWA oly satsfes preferetal depedece wth comootocty domas. We wll show below that our aggregato operators, the geeralzed mxtures wth lear ad quadratc weght geeratg fuctos, do ot satsfy preferetal depedece eve wth comootocty domas. For ths reaso they are able to represet preferece structures ot represetable by OWA operators. The example s as follows. Suppose that we have two ew caddates ad that they oly kow Pascal, Lsp ad Prolog. Table 6 shows the attrbute satsfacto values assocated wth the two ew caddates, usg the same attrbute mportace values as the oes gve Table 4. The smulato cossts creasg the satsfacto value of attrbute Pascal from 0.5 to 0.7. Pascal Lsp Prolog 21

22 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) Caddate D Caddate E Table 6. Ital data for the three attrbutes Pascal, Lsp ad Prolog. The aggregato results are show Table 7. Ideed, a volato of preferetal depedece occurs, ad wth the sgle comootocty doma x < x < x Lsp Pascal Prolog better tha caddate E. : caddate D s tally worse tha caddate E, but eds up Lear OWA Satsfacto Pascal Cad. D Cad. E Cad. D Cad. E Table 7. Aggregato results wth creasg satsfacto of attrbute Pascal ( ) Fgure 5 shows the plot of the aggregato results, llustratg the volato of preferetal depedece produced by the lear weght geeratg fucto. As metoed before, ths meas that our geeralzed mxture operators ca represet preferece structures that are ot possble to represet wth OWA operators, or for that matter wth ay other aggregato operator satsfyg preferetal depedece wth comootocty domas (as, for stace, Choquet tegrals). Dferece D-E Dferece D-E Fgure 5. Preferetal depedece volato wth lear weght geeratg fuctos 22

23 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) Coclusos We have dscussed two geeralzed mxture operators, assocated wth lear ad quadratc weght geeratg fuctos, that pealze poorly satsfed attrbutes ad reward well satsfed attrbutes. We have preseted a example comparg our two geeralzed mxture operators wth the classcal weghted averagg (WA) ad the ordered weghted averagg (OWA) operators. The results obtaed hghlght the potetal ad flexblty of geeralzed mxture operators usg weghtg fuctos that deped o attrbute performaces. Fally, we have show that the geeralzed mxture operators assocated wth (say) lear weght geeratg fuctos do ot satsfy preferetal depedece eve wth comootocty domas ad therefore they ca be used to represet preferece structures whch the OWA operators (ad also the Choquet tegrals) ca ot represet. Refereces Al-Kloub, B., T. Al-Shemmer, et al The role of weghts mult-crtera decso ad, ad the rakg of water projects Jorda. Europea Joural of Operatoal Research 99 : Bellma, R. E. ad L. A. Zadeh Decso-makg a fuzzy evromet. Maagemet Scece 17(4): Che, S.-J. ad C.-L. Hwag Fuzzy multple attrbute decso makg, Sprger-Verlag. Edwards, W. ad J. R. Newma, Eds Multattrbute evaluato. Quattatve applcatos the socal sceces, Sage Publcatos. Grabsch, M Fuzzy tegral multcrtera decso makg. Fuzzy Sets ad Systems 69 : Grabsch, M The applcatos of fuzzy tegrals multcrtera decso makg. Europea Joural of Operatoal Research 89: Hobbs, B. F A comparso of weghtg methods power plat sttg. Decso Sceces 11 : Kaymac, U. ad H. R. va Nauta-Lemke A sestvty aalyss approach to troducg weght factors to decso fucto fuzzy multcrtera decso makg. Fuzzy Sets ad Systems 97: Keeey, R. L. ad H. Raffa Multattrbute prefereces uder ucertaty: more tha two attrbutes. Decsos wth Multple Objectves: Prefereces ad Value Tradeoffs, Joh Wley ad Sos :

24 Draft of paper I: Europea Joural of Operatoal Research 145 (2003) Kckert, W. J. M Fuzzy theores o decso makg. Martus Njhoff Socal Sceces Dvso. Marques Perera, R. A The oress of mxture operators: the expoetal case. Proc. 8th Iteratoal Coferece o Iformato Processg ad Maagemet of Ucertaty Kowledge-Based Systems (IPMU 2000), Madrd, Spa. Marques Perera, R. A. ad G. Pas O o-mootoc aggregato: mxture operators. Proc. 4th Meetg of the EURO Workg Group o Fuzzy Sets (EUROFUSE 99) ad 2d Itl. Coferece o Soft ad Itellget Computg (SIC 99), Budapest, Hugary. Rbero, R. A Fuzzy multple attrbute decso makg: a revew ad ew preferece elctato techques. Fuzzy Sets ad Systems 78: Rbero, R. A Crtera mportace ad satsfacto decso makg. Proc. 8th Iteratoal Coferece o Iformato Processg ad Maagemet of Ucertaty Kowledge-Based Systems (IPMU 2000), Madrd, Spa Rbero, R. A. ad R. A. Marques-Perera Weghts as fuctos of attrbute satsfacto values. Proc. of the Workshop o Preferece Modellg ad Applcatos (EUROFUSE 2001), Graada, Spa. Stllwell, W. G., D. A. Seaver, et al A comparso of weght approxmato techques multattrbute utlty decso makg. Ogazatoal behavour ad Huma Performace 28: Weber, M. ad K. Borcherdg Behavoural flueces o weght judgmets multattrbute decso makg. Europea Joural of Operatoal Research 67: Yager, R. R O ordered weghted averagg aggregato operators multcrtera decso makg. IEEE Trasactos o Systems, Ma, ad Cyberetcs 18(1): Yager, R. R Applcatos ad extesos of OWA aggregatos. It. J. of Ma- Mache Studes 37: Yager, R. R Famles of OWA operators. Fuzzy Sets ad Systems 59: Yager, R. R Quatfer guded aggregato usg OWA operators. It. Joural of Itellget Systems 11: Yager, R. R. ad D. P. Flev Parametrzed AND-lke ad OR-lke OWA operators. It. J. of Geeral Systems 22: Yager, R. R. ad A. Kelma A exteso of the aalytcal herarchy process usg OWA operators. Joural of Itellget & Fuzzy Systems 7(4): Yoo, K The propagato of errors multple-attrbute decso aalyss: a pratcal approach. Joural Operatoal Research Socety 40(7): Yoo, K. P. ad C.-L. Hwag Multple attrbute decso makg, Sage Publcatos. Zadeh, L. A The cocept of a lgustc varable ad ts applcato to approxmate reasog - I. Fuzzy Sets ad Applcatos: Selected Papers by L. A. Zadeh. R. R. Yager, S. Ovchkv, R. Tog ad H. T. Nguye. New York, Joh Wley & Sos. Reprted from: Iformato Sceces, 8 (1975): Zmmerma, H.-J Fuzzy sets, decso makg, ad expert systems. Bosto, Kluwer Academc Publshers. 24

Solving Constrained Flow-Shop Scheduling. Problems with Three Machines

Solving Constrained Flow-Shop Scheduling. Problems with Three Machines It J Cotemp Math Sceces, Vol 5, 2010, o 19, 921-929 Solvg Costraed Flow-Shop Schedulg Problems wth Three Maches P Pada ad P Rajedra Departmet of Mathematcs, School of Advaced Sceces, VIT Uversty, Vellore-632

More information

Functions of Random Variables

Functions of Random Variables Fuctos of Radom Varables Chapter Fve Fuctos of Radom Varables 5. Itroducto A geeral egeerg aalyss model s show Fg. 5.. The model output (respose) cotas the performaces of a system or product, such as weght,

More information

Summary of the lecture in Biostatistics

Summary of the lecture in Biostatistics Summary of the lecture Bostatstcs Probablty Desty Fucto For a cotuos radom varable, a probablty desty fucto s a fucto such that: 0 dx a b) b a dx A probablty desty fucto provdes a smple descrpto of the

More information

An Indian Journal FULL PAPER ABSTRACT KEYWORDS. Trade Science Inc. Research on scheme evaluation method of automation mechatronic systems

An Indian Journal FULL PAPER ABSTRACT KEYWORDS. Trade Science Inc. Research on scheme evaluation method of automation mechatronic systems [ype text] [ype text] [ype text] ISSN : 0974-7435 Volume 0 Issue 6 Boechology 204 Ida Joural FULL PPER BIJ, 0(6, 204 [927-9275] Research o scheme evaluato method of automato mechatroc systems BSRC Che

More information

Analyzing Fuzzy System Reliability Using Vague Set Theory

Analyzing Fuzzy System Reliability Using Vague Set Theory Iteratoal Joural of Appled Scece ad Egeerg 2003., : 82-88 Aalyzg Fuzzy System Relablty sg Vague Set Theory Shy-Mg Che Departmet of Computer Scece ad Iformato Egeerg, Natoal Tawa versty of Scece ad Techology,

More information

Fuzzy TOPSIS Based on α Level Set for Academic Staff Selection

Fuzzy TOPSIS Based on α Level Set for Academic Staff Selection Gadg Busess ad Maagemet Joural Vol. No., 57-70, 007 Fuzzy TOPSIS Based o evel Set for Academc Staff Selecto Nazrah Raml Nor Azzah M. Yacob Faculty of Iformato Techology ad Quattatve Scece Uverst Tekolog

More information

CHAPTER 4 RADICAL EXPRESSIONS

CHAPTER 4 RADICAL EXPRESSIONS 6 CHAPTER RADICAL EXPRESSIONS. The th Root of a Real Number A real umber a s called the th root of a real umber b f Thus, for example: s a square root of sce. s also a square root of sce ( ). s a cube

More information

Ranking Bank Branches with Interval Data By IAHP and TOPSIS

Ranking Bank Branches with Interval Data By IAHP and TOPSIS Rag Ba Braches wth terval Data By HP ad TPSS Tayebeh Rezaetazaa Departmet of Mathematcs, slamc zad Uversty, Badar bbas Brach, Badar bbas, ra Mahaz Barhordarahmad Departmet of Mathematcs, slamc zad Uversty,

More information

Estimation of Stress- Strength Reliability model using finite mixture of exponential distributions

Estimation of Stress- Strength Reliability model using finite mixture of exponential distributions Iteratoal Joural of Computatoal Egeerg Research Vol, 0 Issue, Estmato of Stress- Stregth Relablty model usg fte mxture of expoetal dstrbutos K.Sadhya, T.S.Umamaheswar Departmet of Mathematcs, Lal Bhadur

More information

Part 4b Asymptotic Results for MRR2 using PRESS. Recall that the PRESS statistic is a special type of cross validation procedure (see Allen (1971))

Part 4b Asymptotic Results for MRR2 using PRESS. Recall that the PRESS statistic is a special type of cross validation procedure (see Allen (1971)) art 4b Asymptotc Results for MRR usg RESS Recall that the RESS statstc s a specal type of cross valdato procedure (see Alle (97)) partcular to the regresso problem ad volves fdg Y $,, the estmate at the

More information

L5 Polynomial / Spline Curves

L5 Polynomial / Spline Curves L5 Polyomal / Sple Curves Cotets Coc sectos Polyomal Curves Hermte Curves Bezer Curves B-Sples No-Uform Ratoal B-Sples (NURBS) Mapulato ad Represetato of Curves Types of Curve Equatos Implct: Descrbe a

More information

Generalization of the Dissimilarity Measure of Fuzzy Sets

Generalization of the Dissimilarity Measure of Fuzzy Sets Iteratoal Mathematcal Forum 2 2007 o. 68 3395-3400 Geeralzato of the Dssmlarty Measure of Fuzzy Sets Faramarz Faghh Boformatcs Laboratory Naobotechology Research Ceter vesa Research Isttute CECR Tehra

More information

Some Notes on the Probability Space of Statistical Surveys

Some Notes on the Probability Space of Statistical Surveys Metodološk zvezk, Vol. 7, No., 200, 7-2 ome Notes o the Probablty pace of tatstcal urveys George Petrakos Abstract Ths paper troduces a formal presetato of samplg process usg prcples ad cocepts from Probablty

More information

CHAPTER VI Statistical Analysis of Experimental Data

CHAPTER VI Statistical Analysis of Experimental Data Chapter VI Statstcal Aalyss of Expermetal Data CHAPTER VI Statstcal Aalyss of Expermetal Data Measuremets do ot lead to a uque value. Ths s a result of the multtude of errors (maly radom errors) that ca

More information

TESTS BASED ON MAXIMUM LIKELIHOOD

TESTS BASED ON MAXIMUM LIKELIHOOD ESE 5 Toy E. Smth. The Basc Example. TESTS BASED ON MAXIMUM LIKELIHOOD To llustrate the propertes of maxmum lkelhood estmates ad tests, we cosder the smplest possble case of estmatg the mea of the ormal

More information

Simple Linear Regression

Simple Linear Regression Statstcal Methods I (EST 75) Page 139 Smple Lear Regresso Smple regresso applcatos are used to ft a model descrbg a lear relatoshp betwee two varables. The aspects of least squares regresso ad correlato

More information

Chapter 5 Properties of a Random Sample

Chapter 5 Properties of a Random Sample Lecture 6 o BST 63: Statstcal Theory I Ku Zhag, /0/008 Revew for the prevous lecture Cocepts: t-dstrbuto, F-dstrbuto Theorems: Dstrbutos of sample mea ad sample varace, relatoshp betwee sample mea ad sample

More information

1 Lyapunov Stability Theory

1 Lyapunov Stability Theory Lyapuov Stablty heory I ths secto we cosder proofs of stablty of equlbra of autoomous systems. hs s stadard theory for olear systems, ad oe of the most mportat tools the aalyss of olear systems. It may

More information

COMPROMISE HYPERSPHERE FOR STOCHASTIC DOMINANCE MODEL

COMPROMISE HYPERSPHERE FOR STOCHASTIC DOMINANCE MODEL Sebasta Starz COMPROMISE HYPERSPHERE FOR STOCHASTIC DOMINANCE MODEL Abstract The am of the work s to preset a method of rakg a fte set of dscrete radom varables. The proposed method s based o two approaches:

More information

Ordinary Least Squares Regression. Simple Regression. Algebra and Assumptions.

Ordinary Least Squares Regression. Simple Regression. Algebra and Assumptions. Ordary Least Squares egresso. Smple egresso. Algebra ad Assumptos. I ths part of the course we are gog to study a techque for aalysg the lear relatoshp betwee two varables Y ad X. We have pars of observatos

More information

The Mathematical Appendix

The Mathematical Appendix The Mathematcal Appedx Defto A: If ( Λ, Ω, where ( λ λ λ whch the probablty dstrbutos,,..., Defto A. uppose that ( Λ,,..., s a expermet type, the σ-algebra o λ λ λ are defed s deoted by ( (,,...,, σ Ω.

More information

Management Science Letters

Management Science Letters Maagemet Scece Letters 2 (202) 29 42 Cotets lsts avalable at GrowgScece Maagemet Scece Letters homepage: www.growgscece.com/msl A goal programmg method for dervg fuzzy prortes of crtera from cosstet fuzzy

More information

Cubic Nonpolynomial Spline Approach to the Solution of a Second Order Two-Point Boundary Value Problem

Cubic Nonpolynomial Spline Approach to the Solution of a Second Order Two-Point Boundary Value Problem Joural of Amerca Scece ;6( Cubc Nopolyomal Sple Approach to the Soluto of a Secod Order Two-Pot Boudary Value Problem W.K. Zahra, F.A. Abd El-Salam, A.A. El-Sabbagh ad Z.A. ZAk * Departmet of Egeerg athematcs

More information

Analysis of Lagrange Interpolation Formula

Analysis of Lagrange Interpolation Formula P IJISET - Iteratoal Joural of Iovatve Scece, Egeerg & Techology, Vol. Issue, December 4. www.jset.com ISS 348 7968 Aalyss of Lagrage Iterpolato Formula Vjay Dahya PDepartmet of MathematcsMaharaja Surajmal

More information

Multi Objective Fuzzy Inventory Model with. Demand Dependent Unit Cost and Lead Time. Constraints A Karush Kuhn Tucker Conditions.

Multi Objective Fuzzy Inventory Model with. Demand Dependent Unit Cost and Lead Time. Constraints A Karush Kuhn Tucker Conditions. It. Joural of Math. Aalyss, Vol. 8, 204, o. 4, 87-93 HIKARI Ltd, www.m-hkar.com http://dx.do.org/0.2988/jma.204.30252 Mult Objectve Fuzzy Ivetory Model wth Demad Depedet Ut Cost ad Lead Tme Costrats A

More information

Assignment 5/MATH 247/Winter Due: Friday, February 19 in class (!) (answers will be posted right after class)

Assignment 5/MATH 247/Winter Due: Friday, February 19 in class (!) (answers will be posted right after class) Assgmet 5/MATH 7/Wter 00 Due: Frday, February 9 class (!) (aswers wll be posted rght after class) As usual, there are peces of text, before the questos [], [], themselves. Recall: For the quadratc form

More information

Bayes (Naïve or not) Classifiers: Generative Approach

Bayes (Naïve or not) Classifiers: Generative Approach Logstc regresso Bayes (Naïve or ot) Classfers: Geeratve Approach What do we mea by Geeratve approach: Lear p(y), p(x y) ad the apply bayes rule to compute p(y x) for makg predctos Ths s essetally makg

More information

On Fuzzy Arithmetic, Possibility Theory and Theory of Evidence

On Fuzzy Arithmetic, Possibility Theory and Theory of Evidence O Fuzzy rthmetc, Possblty Theory ad Theory of Evdece suco P. Cucala, Jose Vllar Isttute of Research Techology Uversdad Potfca Comllas C/ Sata Cruz de Marceado 6 8 Madrd. Spa bstract Ths paper explores

More information

COMPARISON OF ANALYTIC HIERARCHY PROCESS AND SOME NEW OPTIMIZATION PROCEDURES FOR RATIO SCALING

COMPARISON OF ANALYTIC HIERARCHY PROCESS AND SOME NEW OPTIMIZATION PROCEDURES FOR RATIO SCALING Please cte ths artcle as: Paweł Kazbudzk, Comparso of aalytc herarchy process ad some ew optmzato procedures for rato scalg, Scetfc Research of the Isttute of Mathematcs ad Computer Scece, 0, Volume 0,

More information

Mu Sequences/Series Solutions National Convention 2014

Mu Sequences/Series Solutions National Convention 2014 Mu Sequeces/Seres Solutos Natoal Coveto 04 C 6 E A 6C A 6 B B 7 A D 7 D C 7 A B 8 A B 8 A C 8 E 4 B 9 B 4 E 9 B 4 C 9 E C 0 A A 0 D B 0 C C Usg basc propertes of arthmetc sequeces, we fd a ad bm m We eed

More information

Block-Based Compact Thermal Modeling of Semiconductor Integrated Circuits

Block-Based Compact Thermal Modeling of Semiconductor Integrated Circuits Block-Based Compact hermal Modelg of Semcoductor Itegrated Crcuts Master s hess Defese Caddate: Jg Ba Commttee Members: Dr. Mg-Cheg Cheg Dr. Daqg Hou Dr. Robert Schllg July 27, 2009 Outle Itroducto Backgroud

More information

Lecture 7. Confidence Intervals and Hypothesis Tests in the Simple CLR Model

Lecture 7. Confidence Intervals and Hypothesis Tests in the Simple CLR Model Lecture 7. Cofdece Itervals ad Hypothess Tests the Smple CLR Model I lecture 6 we troduced the Classcal Lear Regresso (CLR) model that s the radom expermet of whch the data Y,,, K, are the outcomes. The

More information

A New Method for Solving Fuzzy Linear. Programming by Solving Linear Programming

A New Method for Solving Fuzzy Linear. Programming by Solving Linear Programming ppled Matheatcal Sceces Vol 008 o 50 7-80 New Method for Solvg Fuzzy Lear Prograg by Solvg Lear Prograg S H Nasser a Departet of Matheatcs Faculty of Basc Sceces Mazadara Uversty Babolsar Ira b The Research

More information

Chapter 8. Inferences about More Than Two Population Central Values

Chapter 8. Inferences about More Than Two Population Central Values Chapter 8. Ifereces about More Tha Two Populato Cetral Values Case tudy: Effect of Tmg of the Treatmet of Port-We tas wth Lasers ) To vestgate whether treatmet at a youg age would yeld better results tha

More information

Multiple Regression. More than 2 variables! Grade on Final. Multiple Regression 11/21/2012. Exam 2 Grades. Exam 2 Re-grades

Multiple Regression. More than 2 variables! Grade on Final. Multiple Regression 11/21/2012. Exam 2 Grades. Exam 2 Re-grades STAT 101 Dr. Kar Lock Morga 11/20/12 Exam 2 Grades Multple Regresso SECTIONS 9.2, 10.1, 10.2 Multple explaatory varables (10.1) Parttog varablty R 2, ANOVA (9.2) Codtos resdual plot (10.2) Trasformatos

More information

Third handout: On the Gini Index

Third handout: On the Gini Index Thrd hadout: O the dex Corrado, a tala statstca, proposed (, 9, 96) to measure absolute equalt va the mea dfferece whch s defed as ( / ) where refers to the total umber of dvduals socet. Assume that. The

More information

Econometric Methods. Review of Estimation

Econometric Methods. Review of Estimation Ecoometrc Methods Revew of Estmato Estmatg the populato mea Radom samplg Pot ad terval estmators Lear estmators Ubased estmators Lear Ubased Estmators (LUEs) Effcecy (mmum varace) ad Best Lear Ubased Estmators

More information

Lecture 07: Poles and Zeros

Lecture 07: Poles and Zeros Lecture 07: Poles ad Zeros Defto of poles ad zeros The trasfer fucto provdes a bass for determg mportat system respose characterstcs wthout solvg the complete dfferetal equato. As defed, the trasfer fucto

More information

Lecture 3 Probability review (cont d)

Lecture 3 Probability review (cont d) STATS 00: Itroducto to Statstcal Iferece Autum 06 Lecture 3 Probablty revew (cot d) 3. Jot dstrbutos If radom varables X,..., X k are depedet, the ther dstrbuto may be specfed by specfyg the dvdual dstrbuto

More information

MEASURES OF DISPERSION

MEASURES OF DISPERSION MEASURES OF DISPERSION Measure of Cetral Tedecy: Measures of Cetral Tedecy ad Dsperso ) Mathematcal Average: a) Arthmetc mea (A.M.) b) Geometrc mea (G.M.) c) Harmoc mea (H.M.) ) Averages of Posto: a) Meda

More information

CHAPTER 3 POSTERIOR DISTRIBUTIONS

CHAPTER 3 POSTERIOR DISTRIBUTIONS CHAPTER 3 POSTERIOR DISTRIBUTIONS If scece caot measure the degree of probablt volved, so much the worse for scece. The practcal ma wll stck to hs apprecatve methods utl t does, or wll accept the results

More information

It is Advantageous to Make a Syllabus as Precise as Possible: Decision-Theoretic Analysis

It is Advantageous to Make a Syllabus as Precise as Possible: Decision-Theoretic Analysis Joural of Iovatve Techology ad Educato, Vol. 4, 2017, o. 1, 1-5 HIKARI Ltd, www.m-hkar.com https://do.org/10.12988/jte.2017.61146 It s Advatageous to Make a Syllabus as Precse as Possble: Decso-Theoretc

More information

The Necessarily Efficient Point Method for Interval Molp Problems

The Necessarily Efficient Point Method for Interval Molp Problems ISS 6-69 Eglad K Joural of Iformato ad omputg Scece Vol. o. 9 pp. - The ecessarly Effcet Pot Method for Iterval Molp Problems Hassa Mshmast eh ad Marzeh Alezhad + Mathematcs Departmet versty of Ssta ad

More information

Point Estimation: definition of estimators

Point Estimation: definition of estimators Pot Estmato: defto of estmators Pot estmator: ay fucto W (X,..., X ) of a data sample. The exercse of pot estmato s to use partcular fuctos of the data order to estmate certa ukow populato parameters.

More information

ECONOMETRIC THEORY. MODULE VIII Lecture - 26 Heteroskedasticity

ECONOMETRIC THEORY. MODULE VIII Lecture - 26 Heteroskedasticity ECONOMETRIC THEORY MODULE VIII Lecture - 6 Heteroskedastcty Dr. Shalabh Departmet of Mathematcs ad Statstcs Ida Isttute of Techology Kapur . Breusch Paga test Ths test ca be appled whe the replcated data

More information

TWO NEW WEIGHTED MEASURES OF FUZZY ENTROPY AND THEIR PROPERTIES

TWO NEW WEIGHTED MEASURES OF FUZZY ENTROPY AND THEIR PROPERTIES merca. Jr. of Mathematcs ad Sceces Vol., No.,(Jauary 0) Copyrght Md Reader Publcatos www.jouralshub.com TWO NEW WEIGTED MESURES OF FUZZY ENTROPY ND TEIR PROPERTIES R.K.Tul Departmet of Mathematcs S.S.M.

More information

X ε ) = 0, or equivalently, lim

X ε ) = 0, or equivalently, lim Revew for the prevous lecture Cocepts: order statstcs Theorems: Dstrbutos of order statstcs Examples: How to get the dstrbuto of order statstcs Chapter 5 Propertes of a Radom Sample Secto 55 Covergece

More information

Lecture 3. Sampling, sampling distributions, and parameter estimation

Lecture 3. Sampling, sampling distributions, and parameter estimation Lecture 3 Samplg, samplg dstrbutos, ad parameter estmato Samplg Defto Populato s defed as the collecto of all the possble observatos of terest. The collecto of observatos we take from the populato s called

More information

MULTIDIMENSIONAL HETEROGENEOUS VARIABLE PREDICTION BASED ON EXPERTS STATEMENTS. Gennadiy Lbov, Maxim Gerasimov

MULTIDIMENSIONAL HETEROGENEOUS VARIABLE PREDICTION BASED ON EXPERTS STATEMENTS. Gennadiy Lbov, Maxim Gerasimov Iteratoal Boo Seres "Iformato Scece ad Computg" 97 MULTIIMNSIONAL HTROGNOUS VARIABL PRICTION BAS ON PRTS STATMNTS Geady Lbov Maxm Gerasmov Abstract: I the wors [ ] we proposed a approach of formg a cosesus

More information

Chapter 4 Multiple Random Variables

Chapter 4 Multiple Random Variables Revew for the prevous lecture: Theorems ad Examples: How to obta the pmf (pdf) of U = g (, Y) ad V = g (, Y) Chapter 4 Multple Radom Varables Chapter 44 Herarchcal Models ad Mxture Dstrbutos Examples:

More information

UNIT 2 SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

UNIT 2 SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS Numercal Computg -I UNIT SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS Structure Page Nos..0 Itroducto 6. Objectves 7. Ital Approxmato to a Root 7. Bsecto Method 8.. Error Aalyss 9.4 Regula Fals Method

More information

Department of Agricultural Economics. PhD Qualifier Examination. August 2011

Department of Agricultural Economics. PhD Qualifier Examination. August 2011 Departmet of Agrcultural Ecoomcs PhD Qualfer Examato August 0 Istructos: The exam cossts of sx questos You must aswer all questos If you eed a assumpto to complete a questo, state the assumpto clearly

More information

Solution of General Dual Fuzzy Linear Systems. Using ABS Algorithm

Solution of General Dual Fuzzy Linear Systems. Using ABS Algorithm Appled Mathematcal Sceces, Vol 6, 0, o 4, 63-7 Soluto of Geeral Dual Fuzzy Lear Systems Usg ABS Algorthm M A Farborz Aragh * ad M M ossezadeh Departmet of Mathematcs, Islamc Azad Uversty Cetral ehra Brach,

More information

Solving Interval and Fuzzy Multi Objective. Linear Programming Problem. by Necessarily Efficiency Points

Solving Interval and Fuzzy Multi Objective. Linear Programming Problem. by Necessarily Efficiency Points Iteratoal Mathematcal Forum, 3, 2008, o. 3, 99-06 Solvg Iterval ad Fuzzy Mult Obectve ear Programmg Problem by Necessarly Effcecy Pots Hassa Mshmast Neh ad Marzeh Aleghad Mathematcs Departmet, Faculty

More information

7.0 Equality Contraints: Lagrange Multipliers

7.0 Equality Contraints: Lagrange Multipliers Systes Optzato 7.0 Equalty Cotrats: Lagrage Multplers Cosder the zato of a o-lear fucto subject to equalty costrats: g f() R ( ) 0 ( ) (7.) where the g ( ) are possbly also olear fuctos, ad < otherwse

More information

Investigating Cellular Automata

Investigating Cellular Automata Researcher: Taylor Dupuy Advsor: Aaro Wootto Semester: Fall 4 Ivestgatg Cellular Automata A Overvew of Cellular Automata: Cellular Automata are smple computer programs that geerate rows of black ad whte

More information

A New Measure of Probabilistic Entropy. and its Properties

A New Measure of Probabilistic Entropy. and its Properties Appled Mathematcal Sceces, Vol. 4, 200, o. 28, 387-394 A New Measure of Probablstc Etropy ad ts Propertes Rajeesh Kumar Departmet of Mathematcs Kurukshetra Uversty Kurukshetra, Ida rajeesh_kuk@redffmal.com

More information

THE ROYAL STATISTICAL SOCIETY HIGHER CERTIFICATE

THE ROYAL STATISTICAL SOCIETY HIGHER CERTIFICATE THE ROYAL STATISTICAL SOCIETY 00 EXAMINATIONS SOLUTIONS HIGHER CERTIFICATE PAPER I STATISTICAL THEORY The Socety provdes these solutos to assst caddates preparg for the examatos future years ad for the

More information

Application of Calibration Approach for Regression Coefficient Estimation under Two-stage Sampling Design

Application of Calibration Approach for Regression Coefficient Estimation under Two-stage Sampling Design Authors: Pradp Basak, Kaustav Adtya, Hukum Chadra ad U.C. Sud Applcato of Calbrato Approach for Regresso Coeffcet Estmato uder Two-stage Samplg Desg Pradp Basak, Kaustav Adtya, Hukum Chadra ad U.C. Sud

More information

Bayes Estimator for Exponential Distribution with Extension of Jeffery Prior Information

Bayes Estimator for Exponential Distribution with Extension of Jeffery Prior Information Malaysa Joural of Mathematcal Sceces (): 97- (9) Bayes Estmator for Expoetal Dstrbuto wth Exteso of Jeffery Pror Iformato Hadeel Salm Al-Kutub ad Noor Akma Ibrahm Isttute for Mathematcal Research, Uverst

More information

Analysis of Variance with Weibull Data

Analysis of Variance with Weibull Data Aalyss of Varace wth Webull Data Lahaa Watthaacheewaul Abstract I statstcal data aalyss by aalyss of varace, the usual basc assumptos are that the model s addtve ad the errors are radomly, depedetly, ad

More information

Non-uniform Turán-type problems

Non-uniform Turán-type problems Joural of Combatoral Theory, Seres A 111 2005 106 110 wwwelsevercomlocatecta No-uform Turá-type problems DhruvMubay 1, Y Zhao 2 Departmet of Mathematcs, Statstcs, ad Computer Scece, Uversty of Illos at

More information

2.28 The Wall Street Journal is probably referring to the average number of cubes used per glass measured for some population that they have chosen.

2.28 The Wall Street Journal is probably referring to the average number of cubes used per glass measured for some population that they have chosen. .5 x 54.5 a. x 7. 786 7 b. The raked observatos are: 7.4, 7.5, 7.7, 7.8, 7.9, 8.0, 8.. Sce the sample sze 7 s odd, the meda s the (+)/ 4 th raked observato, or meda 7.8 c. The cosumer would more lkely

More information

Chapter 9 Jordan Block Matrices

Chapter 9 Jordan Block Matrices Chapter 9 Jorda Block atrces I ths chapter we wll solve the followg problem. Gve a lear operator T fd a bass R of F such that the matrx R (T) s as smple as possble. f course smple s a matter of taste.

More information

An Introduction to. Support Vector Machine

An Introduction to. Support Vector Machine A Itroducto to Support Vector Mache Support Vector Mache (SVM) A classfer derved from statstcal learg theory by Vapk, et al. 99 SVM became famous whe, usg mages as put, t gave accuracy comparable to eural-etwork

More information

best estimate (mean) for X uncertainty or error in the measurement (systematic, random or statistical) best

best estimate (mean) for X uncertainty or error in the measurement (systematic, random or statistical) best Error Aalyss Preamble Wheever a measuremet s made, the result followg from that measuremet s always subject to ucertaty The ucertaty ca be reduced by makg several measuremets of the same quatty or by mprovg

More information

Ideal multigrades with trigonometric coefficients

Ideal multigrades with trigonometric coefficients Ideal multgrades wth trgoometrc coeffcets Zarathustra Brady December 13, 010 1 The problem A (, k) multgrade s defed as a par of dstct sets of tegers such that (a 1,..., a ; b 1,..., b ) a j = =1 for all

More information

{ }{ ( )} (, ) = ( ) ( ) ( ) Chapter 14 Exercises in Sampling Theory. Exercise 1 (Simple random sampling): Solution:

{ }{ ( )} (, ) = ( ) ( ) ( ) Chapter 14 Exercises in Sampling Theory. Exercise 1 (Simple random sampling): Solution: Chapter 4 Exercses Samplg Theory Exercse (Smple radom samplg: Let there be two correlated radom varables X ad A sample of sze s draw from a populato by smple radom samplg wthout replacemet The observed

More information

Sufficiency in Blackwell s theorem

Sufficiency in Blackwell s theorem Mathematcal Socal Sceces 46 (23) 21 25 www.elsever.com/locate/ecobase Suffcecy Blacwell s theorem Agesza Belsa-Kwapsz* Departmet of Agrcultural Ecoomcs ad Ecoomcs, Motaa State Uversty, Bozema, MT 59717,

More information

MAX-MIN AND MIN-MAX VALUES OF VARIOUS MEASURES OF FUZZY DIVERGENCE

MAX-MIN AND MIN-MAX VALUES OF VARIOUS MEASURES OF FUZZY DIVERGENCE merca Jr of Mathematcs ad Sceces Vol, No,(Jauary 0) Copyrght Md Reader Publcatos wwwjouralshubcom MX-MIN ND MIN-MX VLUES OF VRIOUS MESURES OF FUZZY DIVERGENCE RKTul Departmet of Mathematcs SSM College

More information

Validating Multiattribute Decision Making Methods for Supporting Group Decisions

Validating Multiattribute Decision Making Methods for Supporting Group Decisions Valdatg Multattrbute Decso Makg Methods for Supportg Group Decsos Chug-Hsg Yeh, Seor Member, IEEE Clayto School of Iformato Techology Faculty of Iformato Techology, Moash Uversty Clayto, Vctora, 3800,

More information

Statistics: Unlocking the Power of Data Lock 5

Statistics: Unlocking the Power of Data Lock 5 STAT 0 Dr. Kar Lock Morga Exam 2 Grades: I- Class Multple Regresso SECTIONS 9.2, 0., 0.2 Multple explaatory varables (0.) Parttog varablty R 2, ANOVA (9.2) Codtos resdual plot (0.2) Exam 2 Re- grades Re-

More information

1. BLAST (Karlin Altschul) Statistics

1. BLAST (Karlin Altschul) Statistics Parwse seuece algmet global ad local Multple seuece algmet Substtuto matrces Database searchg global local BLAST Seuece statstcs Evolutoary tree recostructo Gee Fdg Prote structure predcto RNA structure

More information

Bayesian Classification. CS690L Data Mining: Classification(2) Bayesian Theorem: Basics. Bayesian Theorem. Training dataset. Naïve Bayes Classifier

Bayesian Classification. CS690L Data Mining: Classification(2) Bayesian Theorem: Basics. Bayesian Theorem. Training dataset. Naïve Bayes Classifier Baa Classfcato CS6L Data Mg: Classfcato() Referece: J. Ha ad M. Kamber, Data Mg: Cocepts ad Techques robablstc learg: Calculate explct probabltes for hypothess, amog the most practcal approaches to certa

More information

Module 7: Probability and Statistics

Module 7: Probability and Statistics Lecture 4: Goodess of ft tests. Itroducto Module 7: Probablty ad Statstcs I the prevous two lectures, the cocepts, steps ad applcatos of Hypotheses testg were dscussed. Hypotheses testg may be used to

More information

A New Family of Transformations for Lifetime Data

A New Family of Transformations for Lifetime Data Proceedgs of the World Cogress o Egeerg 4 Vol I, WCE 4, July - 4, 4, Lodo, U.K. A New Famly of Trasformatos for Lfetme Data Lakhaa Watthaacheewakul Abstract A famly of trasformatos s the oe of several

More information

. The set of these sums. be a partition of [ ab, ]. Consider the sum f( x) f( x 1)

. The set of these sums. be a partition of [ ab, ]. Consider the sum f( x) f( x 1) Chapter 7 Fuctos o Bouded Varato. Subject: Real Aalyss Level: M.Sc. Source: Syed Gul Shah (Charma, Departmet o Mathematcs, US Sargodha Collected & Composed by: Atq ur Rehma (atq@mathcty.org, http://www.mathcty.org

More information

Comparing Different Estimators of three Parameters for Transmuted Weibull Distribution

Comparing Different Estimators of three Parameters for Transmuted Weibull Distribution Global Joural of Pure ad Appled Mathematcs. ISSN 0973-768 Volume 3, Number 9 (207), pp. 55-528 Research Ida Publcatos http://www.rpublcato.com Comparg Dfferet Estmators of three Parameters for Trasmuted

More information

DIFFERENTIAL GEOMETRIC APPROACH TO HAMILTONIAN MECHANICS

DIFFERENTIAL GEOMETRIC APPROACH TO HAMILTONIAN MECHANICS DIFFERENTIAL GEOMETRIC APPROACH TO HAMILTONIAN MECHANICS Course Project: Classcal Mechacs (PHY 40) Suja Dabholkar (Y430) Sul Yeshwath (Y444). Itroducto Hamltoa mechacs s geometry phase space. It deals

More information

MATH 247/Winter Notes on the adjoint and on normal operators.

MATH 247/Winter Notes on the adjoint and on normal operators. MATH 47/Wter 00 Notes o the adjot ad o ormal operators I these otes, V s a fte dmesoal er product space over, wth gve er * product uv, T, S, T, are lear operators o V U, W are subspaces of V Whe we say

More information

AN UPPER BOUND FOR THE PERMANENT VERSUS DETERMINANT PROBLEM BRUNO GRENET

AN UPPER BOUND FOR THE PERMANENT VERSUS DETERMINANT PROBLEM BRUNO GRENET AN UPPER BOUND FOR THE PERMANENT VERSUS DETERMINANT PROBLEM BRUNO GRENET Abstract. The Permaet versus Determat problem s the followg: Gve a matrx X of determates over a feld of characterstc dfferet from

More information

LINEAR REGRESSION ANALYSIS

LINEAR REGRESSION ANALYSIS LINEAR REGRESSION ANALYSIS MODULE V Lecture - Correctg Model Iadequaces Through Trasformato ad Weghtg Dr. Shalabh Departmet of Mathematcs ad Statstcs Ida Isttute of Techology Kapur Aalytcal methods for

More information

Chapter 14 Logistic Regression Models

Chapter 14 Logistic Regression Models Chapter 4 Logstc Regresso Models I the lear regresso model X β + ε, there are two types of varables explaatory varables X, X,, X k ad study varable y These varables ca be measured o a cotuous scale as

More information

Objectives of Multiple Regression

Objectives of Multiple Regression Obectves of Multple Regresso Establsh the lear equato that best predcts values of a depedet varable Y usg more tha oe eplaator varable from a large set of potetal predctors {,,... k }. Fd that subset of

More information

ENGI 3423 Simple Linear Regression Page 12-01

ENGI 3423 Simple Linear Regression Page 12-01 ENGI 343 mple Lear Regresso Page - mple Lear Regresso ometmes a expermet s set up where the expermeter has cotrol over the values of oe or more varables X ad measures the resultg values of aother varable

More information

TRIANGULAR MEMBERSHIP FUNCTIONS FOR SOLVING SINGLE AND MULTIOBJECTIVE FUZZY LINEAR PROGRAMMING PROBLEM.

TRIANGULAR MEMBERSHIP FUNCTIONS FOR SOLVING SINGLE AND MULTIOBJECTIVE FUZZY LINEAR PROGRAMMING PROBLEM. Abbas Iraq Joural of SceceVol 53No 12012 Pp. 125-129 TRIANGULAR MEMBERSHIP FUNCTIONS FOR SOLVING SINGLE AND MULTIOBJECTIVE FUZZY LINEAR PROGRAMMING PROBLEM. Iraq Tarq Abbas Departemet of Mathematc College

More information

Lecture 1 Review of Fundamental Statistical Concepts

Lecture 1 Review of Fundamental Statistical Concepts Lecture Revew of Fudametal Statstcal Cocepts Measures of Cetral Tedecy ad Dsperso A word about otato for ths class: Idvduals a populato are desgated, where the dex rages from to N, ad N s the total umber

More information

Group decision-making based on heterogeneous preference. relations with self-confidence

Group decision-making based on heterogeneous preference. relations with self-confidence Group decso-mag based o heterogeeous preferece relatos wth self-cofdece Yucheg Dog,Weq Lu, Busess School, Schua Uversty, Chegdu 60065, Cha E-mal: ycdog@scu.edu.c; wqlu@stu.scu.edu.c Fracsco Chclaa, Faculty

More information

CS286.2 Lecture 4: Dinur s Proof of the PCP Theorem

CS286.2 Lecture 4: Dinur s Proof of the PCP Theorem CS86. Lecture 4: Dur s Proof of the PCP Theorem Scrbe: Thom Bohdaowcz Prevously, we have prove a weak verso of the PCP theorem: NP PCP 1,1/ (r = poly, q = O(1)). Wth ths result we have the desred costat

More information

Laboratory I.10 It All Adds Up

Laboratory I.10 It All Adds Up Laboratory I. It All Adds Up Goals The studet wll work wth Rema sums ad evaluate them usg Derve. The studet wll see applcatos of tegrals as accumulatos of chages. The studet wll revew curve fttg sklls.

More information

Unimodality Tests for Global Optimization of Single Variable Functions Using Statistical Methods

Unimodality Tests for Global Optimization of Single Variable Functions Using Statistical Methods Malaysa Umodalty Joural Tests of Mathematcal for Global Optmzato Sceces (): of 05 Sgle - 5 Varable (007) Fuctos Usg Statstcal Methods Umodalty Tests for Global Optmzato of Sgle Varable Fuctos Usg Statstcal

More information

BERNSTEIN COLLOCATION METHOD FOR SOLVING NONLINEAR DIFFERENTIAL EQUATIONS. Aysegul Akyuz Dascioglu and Nese Isler

BERNSTEIN COLLOCATION METHOD FOR SOLVING NONLINEAR DIFFERENTIAL EQUATIONS. Aysegul Akyuz Dascioglu and Nese Isler Mathematcal ad Computatoal Applcatos, Vol. 8, No. 3, pp. 293-300, 203 BERNSTEIN COLLOCATION METHOD FOR SOLVING NONLINEAR DIFFERENTIAL EQUATIONS Aysegul Ayuz Dascoglu ad Nese Isler Departmet of Mathematcs,

More information

1 Mixed Quantum State. 2 Density Matrix. CS Density Matrices, von Neumann Entropy 3/7/07 Spring 2007 Lecture 13. ψ = α x x. ρ = p i ψ i ψ i.

1 Mixed Quantum State. 2 Density Matrix. CS Density Matrices, von Neumann Entropy 3/7/07 Spring 2007 Lecture 13. ψ = α x x. ρ = p i ψ i ψ i. CS 94- Desty Matrces, vo Neuma Etropy 3/7/07 Sprg 007 Lecture 3 I ths lecture, we wll dscuss the bascs of quatum formato theory I partcular, we wll dscuss mxed quatum states, desty matrces, vo Neuma etropy

More information

1. A real number x is represented approximately by , and we are told that the relative error is 0.1 %. What is x? Note: There are two answers.

1. A real number x is represented approximately by , and we are told that the relative error is 0.1 %. What is x? Note: There are two answers. PROBLEMS A real umber s represeted appromately by 63, ad we are told that the relatve error s % What s? Note: There are two aswers Ht : Recall that % relatve error s What s the relatve error volved roudg

More information

PROJECTION PROBLEM FOR REGULAR POLYGONS

PROJECTION PROBLEM FOR REGULAR POLYGONS Joural of Mathematcal Sceces: Advaces ad Applcatos Volume, Number, 008, Pages 95-50 PROJECTION PROBLEM FOR REGULAR POLYGONS College of Scece Bejg Forestry Uversty Bejg 0008 P. R. Cha e-mal: sl@bjfu.edu.c

More information

A Study on Generalized Generalized Quasi hyperbolic Kac Moody algebra QHGGH of rank 10

A Study on Generalized Generalized Quasi hyperbolic Kac Moody algebra QHGGH of rank 10 Global Joural of Mathematcal Sceces: Theory ad Practcal. ISSN 974-3 Volume 9, Number 3 (7), pp. 43-4 Iteratoal Research Publcato House http://www.rphouse.com A Study o Geeralzed Geeralzed Quas (9) hyperbolc

More information

X X X E[ ] E X E X. is the ()m n where the ( i,)th. j element is the mean of the ( i,)th., then

X X X E[ ] E X E X. is the ()m n where the ( i,)th. j element is the mean of the ( i,)th., then Secto 5 Vectors of Radom Varables Whe workg wth several radom varables,,..., to arrage them vector form x, t s ofte coveet We ca the make use of matrx algebra to help us orgaze ad mapulate large umbers

More information

Chapter 2 - Free Vibration of Multi-Degree-of-Freedom Systems - II

Chapter 2 - Free Vibration of Multi-Degree-of-Freedom Systems - II CEE49b Chapter - Free Vbrato of Mult-Degree-of-Freedom Systems - II We ca obta a approxmate soluto to the fudametal atural frequecy through a approxmate formula developed usg eergy prcples by Lord Raylegh

More information

Arithmetic Mean and Geometric Mean

Arithmetic Mean and Geometric Mean Acta Mathematca Ntresa Vol, No, p 43 48 ISSN 453-6083 Arthmetc Mea ad Geometrc Mea Mare Varga a * Peter Mchalča b a Departmet of Mathematcs, Faculty of Natural Sceces, Costate the Phlosopher Uversty Ntra,

More information

Entropy ISSN by MDPI

Entropy ISSN by MDPI Etropy 2003, 5, 233-238 Etropy ISSN 1099-4300 2003 by MDPI www.mdp.org/etropy O the Measure Etropy of Addtve Cellular Automata Hasa Aı Arts ad Sceces Faculty, Departmet of Mathematcs, Harra Uversty; 63100,

More information