AN EFFICIENT METHOD BASED ON ABC FOR OPTIMAL MULTILEVEL THRESHOLDING *

Size: px
Start display at page:

Download "AN EFFICIENT METHOD BASED ON ABC FOR OPTIMAL MULTILEVEL THRESHOLDING *"

Transcription

1 IJST, Transactons of Electrcal Engneerng, Vol. 6, No. E, pp 7-9 Prnted n The Islamc Republc of Iran, 0 Shraz Unversty AN EFFICIENT METHOD BASED ON ABC FOR OPTIMAL MULTILEVEL THRESHOLDING * S. A. MOHAMMADI, R. AKBARI ** AND S. H. MOHAMMADI Dept. of Electrcal and Computer Engneerng, Tabrz Unversty, Tabrz, I. R. of Iran Dept. of Computer Engneerng and Informaton Technology, Shraz Unversty of Technology, Shraz, I. R. of Iran Emal: akbar@sutech.ac.r Center for Spoken Language Understandng, Oregon Health & Scence Unversty, Portland, OR, USA Abstract Many effcent b-level thresholdng technques have been proposed n recent years. Usually, the objectve functons, whch are used by them, are not approprate for the multlevel thresholdng owng to exponental growth of computatonal complexty. Ths work presents a new multlevel thresholdng algorthm usng Artfcal Bee Colony algorthm (ABC) wth the Otsu s objectve functon. Also, a strategy s used to guess sutable thresholds for ntalzng the proposed. Ths ntalzng phase used the b-level Otsu to fnd the ntal thresholds. These guessed thresholds are used to create a food source around each of them for use n the ABC algorthm as ntal populaton. The presented thresholdng s tested on four popular mages. The results show that ths has compettve performance compared to other wellknown s such as Gaussan-smoothng, Symmetry-dualty, GA-based and PSO-based algorthms. Keywords Segmentaton, multlevel thresholdng, otsu, ABC algorthm. INTRODUCTION Image segmentaton s one of the most fundamental processes for mage analyss and explanaton. Ths process s used for dstnct objects n an mage from ts background or dstngushng an object from other objects wth dfferent gray-levels. One of the s used broadly n mage segmentaton s mage thresholdng. It can be categorzed nto two classes whch are called b-level thresholdng and multlevel thresholdng. B-level thresholdng parttons pxels nto two dstnct classes; one of the classes comprses those pxels havng gray level values above a predefned threshold and the other class ncludes the rest of the pxels. Multlevel thresholdng splts the pxels nto varous classes and the pxels wthn a class have a gray level nsde a partcular range. In ths, a sngle gray-level value s assgned to each class. One effectve and benefcal tool used for evoluton of mage segmentaton s mage gray level hstograms. Our purpose s to fnd thresholds, whch are gray level values, to classfy pxels. The smplest classfcaton s b-level thresholdng n whch one threshold should be found, but the problem becomes more complcated f more thresholds need to be found. Generally, t s not straghtforward to guess thresholds n the multmodal hstogram of an mage for mage segmentaton. Thus multlevel thresholdng s an nterestng and challengng area of research n the feld of mage processng. Many thresholdng technques have been proposed n the past years [-]. These technques can be classfed nto two dstnct groups: optmal thresholdng s [-8] and property-based thresholdng s [9-]. Optmal thresholdng s try to fnd an optmal threshold that segments classes to acheve approprated attrbutes. Conventonally, ths process s done wth optmzng an objectve functon. Property-based procedures evaluate some propertes of hstogram for detectng the thresholds. Receved by the edtors Aprl, 0; Accepted June 0, 0. Correspondng author

2 8 S. A. Mohammad et al. In recent years many algorthms have been proposed for optmal thresholdng, such as maxmzng a posteror entropy to measure the homogenety of the threshold classes [-7], an optmal thresholdng to maxmze the separateness measures of the classes based on dscrmnant analyss, whch was proposed by Otsu 8] or a mnmum error to mnmze the pxel classfcaton error rate proposed by Kttler and Illngworth [ ]. Because most of these algorthms were orgnally developed for b-level thresholdng, they encounter the computatonal complexty, whch s exponental when extended to multlevel thresholdng due to the exhaustve and comprehensve search. An teratve was proposed to overcome ths problem n [ ]. It assumed the multlevel thresholdng problem as successve b-level thresholdng problems. Then the threshold values were changed to optmze an objectve functon. Ths teratve scheme repeated none of the changed threshold values or the threshold could not move. Unfortunately, ths traps local optma wth three or more thresholds. There are also some proposed s for property-based thresholdng. For nstance, a whch frst smooths the gray-level hstogram by the Gaussan convoluton has been developed by Lm and Lee [ ]. Then the frst and second dervatves of the smoothed hstogram are used to detect thresholds. Another whch has been proposed by Tsa 0] works wth unmodal hstograms by ncludng a local maxmum curvature to recognze the thresholds. Property-based thresholdng s are rapd and approprate for the case of multlevel thresholdng, whle the number of thresholds s dffcult to determne and should be gven beforehand. Karaboga has represented artfcal Bee Colony (ABC) algorthm, whch s based on the foragng behavor of honeybees [ ]. ABC s one of the most recently ntroduced optmzaton algorthms. The prevous studes on ABC showed that ths algorthm s effcent and compettve performance may be obtaned by usng ths algorthm compared to the other algorthms n many engneerng felds. The performance of the ABC algorthm on multlevel thresholdng s compared wth the results of the Partcle Swarm Optmzaton (PSO) algorthm on the same benchmarks that are presented n [ ] and some other multlevel thresholdng algorthms. ABC and PSO algorthms are populaton-based optmzaton algorthms and they have been nspred by the socal behavor of anmals and swarm ntellgence, whch classfy them n the same class of swarm-based optmzaton algorthms. There are many other optmzaton algorthms proposed n lterature such as [ - 7] whch are used n other felds such as thermal unts 8], dscrete optmzaton [ 6] or contnuous optmzaton [ 7]. In ths work, a new algorthm based on ABC algorthm s proposed whch represents compettve results compared to the prevous algorthms. The proposed algorthm uses Otsu mult-level algorthm to ntalze thresholds snce t s very smple and the nearly optmal thresholds are selected based not only on local propertes such as valleys, but also based on the global propertes of the hstogram. However, t falls n the local mnma. Hence, for solvng ths problem, ABC s used as an algorthm to take t back from local mnma and perform a global search. The rest of the paper s organzed as follows: the Otsu s presented n secton. An teratve ntalzng s presented n secton. Proposed ABC multlevel thresholdng algorthm s ntroduced n secton. The expermental results are proposed n secton. Fnally secton 6 concludes ths work.. THE OTSU METHOD Otsu proposed a whch tres to maxmze the between class varance to dsjont the segmented classes as far as possble to detect the optmal threshold for segmentaton. Ths between class varance measure s used as the objectve functon n the ABC algorthm. It s formulated and defned as follows. Assume that an mage contans N pxels n whch the range of gray levels of a gven mage s located n the range of [0, L-] and h() ndcates the occurrence of gray level. We defne N as total number of pxels and P as the probablty of occurrences of gray level n the mage as follows: IJST, Transactons of Electrcal Engneerng, Volume 6, Number E June 0

3 An effcent based on ABC for 9 L N h( ),0 L () 0 h( ) P,0 L () N As a means to evaluate qualty of threshold (at level t) we shall brng between-class varance to the table. We should maxmze f(t) to fnd optmal threshold: where: f ( t) w w ( ) w w () t P () 0 N t P () P t (6) 0 w0 P N (7) t w Suppose we dvded the pxels nto two dstnct parts by a threshold t. The probabltes of each part occurrence are ndcated by w 0 and w and the mean of each part s denoted by μ 0 and μ. We could also wrte Eq. () n the other form as: where: f ( t) w w (8) 0 0 P (9) t 0 ( 0) 0 w0 P N ( ) (0) t w P N ( ) () 0 w N w P () 0 P N () 0 w The σ 0 and σ show varance of each class or part and σ denotes varance over all pxels. The w and the μ respectvely show the probabltes of all pxels occurrence and the mean level of all pxels. Thus, maxmzng Eq. 8 s dentcal to mnmzng: g( t) w w () 0 0 The g(t) functon s also known as wthn-class varance that should be maxmzed. Expandng ths logc to multlevel thresholdng s straghtforward as a result of dscrmnant crteron: f ( t) w w... w () 0 0 c c June 0 IJST, Transactons of Electrcal Engneerng, Volume 6, Number E

4 0 S. A. Mohammad et al. where c s the number of thresholds. Eq. () s used as an objectve functon for the ABC algorthm that should be mnmzed.. AN ITERATIVE INITIALIZING METHOD Otsu could fnd the optmal b-level threshold effectvely. Otsu also extends hs for the multlevel thresholdng but t suffers from computatonal complexty and tme, owng to the comprehensve search. Assume that we want to fnd c optmal thresholds. Due to an exhaustve search for fndng every possble set of thresholds, we obtan complexty of O(n c ) whch grows exponentally as the number of thresholds ncrease. Hence, these s are not approprate for use n practcal applcatons. An teratve s used for utlzng b-level Otsu to fnd several thresholds [ ]. If ths teratve scheme s used n the ntalzng phase of the proposed algorthm, notable computatonal tme n fndng multlevel thresholds s retaned due to lmtng the search space. Ths uses unformty measure as a crteron for evaluatng the qualty of the segmented mage whch s descrbed n secton.. At frst, the teratve procedure uses Otsu b-level thresholdng. Afterwards, to obtan hgher-level thresholds, Otsu b-level thresholdng s used n each sutable segmented class to maxmze the crteron functon (.e. unformty measure) as much as practcable. The pseudo-code of the teratve ntalzng algorthm s gven n Fg.. Apply b-level Otsu s to the whole hstogram. For n= to C Apply Otsu s n every segmented class separately. Compute the value of objectve functon n each case. Determne the class whch produced hghest value of the objectve functon. Store the new threshold value. End Fg.. The pseudocode of the teratve ntalzng ] The complexty of ths algorthm s the same as b-level thresholdng because all values n the range of [0, L-] are examned for every new threshold. It means that b-level thresholdng algorthms are used for c tmes, f we want to fnd c thresholds. Thus the tme complexty of the algorthm s O(c.n), whch s a lnear complexty. Table compares the algorthm runtme wth Gaussan-smoothng, Symmetrydualty, GA learnng-otsu and PSO based wth c = to for Lena and Pepper mages. As can be seen from Table, the CPU-tme results of teratve ntalzng are lnear and quck. Table. Tme complextes of the proposed teratve ntalzng, Gaussan-smoothng, Symmetry-dualty and GA-based Images Lena Peppers c Intal proposed thresholds 7 7,67 76,7,67 76,7,,67 76,7,,67,9 7 0,7 0,7,9,0,7,9,0,,7,9 Intal proposed Gaussan smoothng CPU-tme(s) Symmetry dualty GA-based IJST, Transactons of Electrcal Engneerng, Volume 6, Number E June 0

5 An effcent based on ABC for The unformty values are also computed n Table. The results ndcate that ths ntalzng s superor to other s. From these results one can see that mult-level thresholdng gves a better result n nearly no tme, so t can be consdered a good choce for ntalzng thresholds. Table. Unformty values of the proposed teratve ntalzng, Gaussan-smoothng, Symmetry-dualty and GA-based. Images Lena Peppers c Intal proposed thresholds 7 7,67 76,7,67 76,7,,67 76,7,,67,9 7 0,7 0,7,9,0,7,9,0,,7,9 Unformty values Intal Gaussan proposed smoothng Symmetry dualty GA-based PROPOSED ABC MULTILEVEL THRESHOLDING ALGORITHM Artfcal Bee Colony (ABC) algorthm was developed by Karaboga for numercal optmzaton [ ]. The nspraton behnd ths algorthm s the foragng behavor of honey bee swarms n fndng foods. Some popular and well-known evolutonary algorthms such as Partcle Swarm Optmzaton (PSO) and Genetc Algorthm (GA) have been compared wth the ABC algorthm on some unconstraned and constraned problems [-]. The ABC algorthm s also appled on the clusterng algorthms [ 6] and the tranng of neural networks [7, 8] wth the proposed objectve functon n 9]. The prevous studes showed that ABC outperforms other meta-heurstc algorthms on most of the engneerng problems. ABC algorthm dvdes artfcal bees nto three groups of bees: employed, onlooker and scout bees. Employed bees brng ther nformaton about food sources for onlooker bees. Ths nformaton s prepared for onlooker bees by dancng on dance area. From these dances, the onlookers could realze whch source s better. Another type of bee whch search the area stochastcally, s the scout bee. The artfcal bee colony s dvded nto equal parts. The frst part ncludes the employed bees and the other part conssts of the onlooker bees. If a food source s dscarded by other bees, the employed bee whch found that food source wll turn nto a scout bee. The pseudo-code of the ABC algorthm s presented n Fg.. Frst, P sze food sources or solutons are ntalzed around the results whch are found by the teratve n the preprocessng secton accordng to the followng constrant, that the numbers should be ascendng to not havng overlap between segmented classes. Then employed and onlooker bees are produced where the number of bees n each group s equal to the food sources. Each food source n ths algorthm ndcates a soluton for the optmzaton problem. The amount of nectar or the qualty measure known as ftness s computed as follows: ft (6) f where f s the Otsu s multlevel objectve functon whch s presented n Eq. (). June 0 IJST, Transactons of Electrcal Engneerng, Volume 6, Number E

6 S. A. Mohammad et al. Intalzaton: Generate a random populaton around the food sources suggested by teratve locaton. Z,=,,P sze //P sze denotes the sze of populaton Compute ftness of the populaton as defned by (6) Do For each employed bee Compute soluton v accordng to (7). Compute the value of f Update food source f the new food source s more relable. End Compute probablty of food selecton as specfed by (8). For every onlooker bee Select a soluton accordng to ts probablty. Compute soluton v accordng to (9). Compute the value of f Update food source f the new food source s more relable. End If there s an abandoned food source for the scout Replace soluton wth a new one as specfed by (0). Save the best soluton up to now. Untl termnaton condton s met. Fg.. The pseudocode of the proposed ABC algorthm for thresholdng The employed bees use ther prevous nformaton and ther current local nformaton around them to update ther locatons f the qualty of the found nectar s hgher than ther ex-locatons. Flyng Pattern of the employed bees s computed as follows from what s remembered before: v z ( z z ) (7) j j j j kj where j {,,...,D} and k {,,..., P sze } denote ndces whch are chosen arbtrarly. Coeffcent j s chosen randomly between [, ]. The ftness of v j s then compared wth the prevous food source ftness and f the recent one s better or equal to t, t s exchanged wth the old one. The employed bees return to ther hve to share ther nformaton wth the onlooker bees by some dancng styles, whch demonstrate the drecton and the amount of nectar n the food sources. The onlooker bees evaluate ths nformaton and choose to go to a food source accordng to some probablty functons defned as follows: p j Psze n ft ft n (8) where P sze s the number of food sources whch s equal to the number of both employed bees and onlooker bees, and ft s gven n Eq. (6). The onlooker bees also used ther memorzed best food poston and the newly chosen food source to update ther poston f the qualty of nectar s better than the prevous memorzed poston. The onlooker bees flyng trajectores are updated as follow: v z ( z z ) (9) j j j j wj where j {,,...,D} denote an ndex whch s chosen arbtrarly and w {,,...,P sze } denote an ndex whch s chosen wth probablty that s computed wth formula. Coeffcent j s chosen randomly IJST, Transactons of Electrcal Engneerng, Volume 6, Number E June 0

7 An effcent based on ABC for between [, ]. The ftness of v j s then compared wth the prevous food source ftness and f the new one s better or equal to t, t s replaced wth the old one. The scout bees change a deserted food source wth a haphazardly produced food source. If a food source poston does not change after a predefned number of teraton called lmt, t s assumed as deserted food source. If we presume that the deserted food source s z then the update rule s defned as follows: z z rand(0,)( z z ) (0) j j j j mn max mn where j {,,..., D}. In ABC algorthm, the exploraton s exerted by the employees and the onlooker n the search area and the scout bees n the search space do the explotaton.. EXPERIMENTAL RESULTS Ths secton presents the conducted experments to evaluate the performance of the proposed algorthm on the well-known benchmarks. a) Performance metrc The unformty measure was proposed by Levne and Nazf 9] for measurng mage segmentaton qualty. Next, Sahoo et al. [ ] recommended usng ths measure to test the performance of thresholdng algorthms. Ths qualty measure corresponds to Otsu crteron measure [ 0]. The unformty measure can be formulated as follows: j0 R j ( f ) U c N ( f f ) c where c denotes number of thresholds, R j ndcates the segmented regon j, f represents the gray level of pxel, μ j ponts to the mean of the gray levels of pxels n the segmented regon j, N ndcates the total number of pxels n the gven mage, f max represents the maxmum gray level of pxels n the gven mage, f mn demonstrates the mnmum gray level of pxels n the gven mage. The unformty value s between 0 and. The hgher values whch are closer to means that the qualty of thresholdng s better. b) Settngs of the algorthms The performance of the proposed algorthm n comparson wth the performances of Genetc Algorthm based multple thresholdng [ 0], Gaussan-smoothng [0, ] and Symmetry-dualty [9, ] are presented n ths secton. Four well-known mages (.e. Lena, Peppers, House, and Elane) are used here for comparson wth other s [ ]. The benchmark mages downloaded from ] and [ ] are of sze. c) Parameters tunng It seems that the proposed algorthm provdes effcency n multlevel thresholdng. However, ts performance may be affected by some parameters. The number of teratons and the sze of the colony are two man parameters whch nfluence the performance of the proposed algorthms. Two experments are conducted to observe the behavor of the proposed algorthm under dfferent confguratons. max j mn () June 0 IJST, Transactons of Electrcal Engneerng, Volume 6, Number E

8 S. A. Mohammad et al. In the frst experment, the effect of the colony sze s studed. The sze of colony vares between 0 and 0 n steps of 0. In ths experment, parameter c s set at and the number of teratons s fxed at 0. All the results are the average of 0 ndependent runs. Table shows the performance of the proposed algorthm wth a colony of dfferent szes. From Table, t can be seen that the sze of the colony affects the performance of the proposed algorthm. It seems that the unformty measures for a specfc colony sze of 0 could be the best choce wth dfferent number of teratons. Table. Unformty values for c = and dverse colony sze for four benchmark mages Iteraton 0 C Colony sze Lena Unformty values Peppers House Elane In the second experment, the effect of the teraton numbers s studed. The number of teratons vares between 0 and 0 n steps of 0. In ths experment, parameter c s set equal to and the sze of the colony s fxed at 0. All the results are the average of 0 ndependent runs. Table shows the performance of the proposed algorthm under dfferent number of teratons. From Table, t can be seen that the number of teratons affects the performance of the proposed algorthm. The best results are obtaned when the number of teratons s set at 0. Table. Unformty values for c = and dverse teratons for four benchmark mages Colony sze 0 C Iteratons Lena Unformty values Peppers House Elane These results show that ABC algorthm has better performance wth 0 teratons and a colony sze wth 0 bees. We choose these parameters not only because of slght changes n the results, but the less computaton tme that we gan. Ths could make our proposed algorthm faster than the PSO based snce 00 teratons were used to converge to the optmum soluton wth partcles. Coeffcents dd not affect the performance of the algorthm snce they were unformly generated n a fxed range. d) Comparatve study The performance of the proposed ABC multlevel thresholdng algorthm s compared wth the performance of PSO based, Gaussan-smoothng, GA-learnng-Otsu algorthm and Symmetry-dualty. The unformty values of the proposed algorthm n comparson wth other algorthms are shown n Table. The results show that the proposed s effcent and outperforms other s on most of the experments. The best result on Lena mage for c= s obtaned by the PSO based. Except for ths case, the best results on all other cases were obtaned by the proposed. IJST, Transactons of Electrcal Engneerng, Volume 6, Number E June 0

9 An effcent based on ABC for Table.Unformty values of the proposed ABC multlevel thresholdng, PSO based, Gaussan-smoothng, Symmetry-dualty and GA-based for Lena and Peppers mages Images Lena Peppers c Proposed thresholds 96,6 80,6,70 7,,7,79 7,06,6,9,9 0,9 68,,0 6,7,6,,0,,7,0 Proposed Unformty Values PSO based Gaussan Symmetry smoothng dualty GA-based The proposed s also compared wth basc PSO and adaptve PSO [] for Elane and House mages. Table 6 presents the unformty values of the proposed algorthm n comparson wth PSO and adaptve PSO. The results show that the proposed has compettve performance compared to the other s. The best results on the House mage for c= and c= are obtaned by Adaptve PSO. Except for these cases, the proposed performs better on other cases. As mentoned before, the proposed algorthm s appled for Lena, Peppers, House and Elane mages and dfferent thresholds are used to segment them. For other algorthms, the mentoned thresholds n ther papers are used to represent segmented mages. The results of hybrd GA, PSO and ABC, whch are derved thresholds on the hstogram of mages and ther correspondng mages wth thresholds are shown n Fgs. -6. In the prevous works lke Gaussan-smoothng and symmetry/dualty s, t can be seen that these s tend to put threshold at the valley of hstograms, whch s the best for mages wth dstnct foreground and background but populaton-based algorthms lke PSO or GA can work wth more complex hstograms whch may have unequal-sze peaks or flat valleys [ ]. These s care more about the characterstcs of mages based on dscrmnant analyss, whch can be seen n Fgs. -6. As t can be seen n Fg. b, Fg. b, Fg. b the qualty of mages wth thresholds s better than other mages, but the qualty of Fg. 6b and Fg. 6c are nearly the same as each other snce the founded thresholds are nearly smlar. Table 6.Unformty values of the proposed ABC multlevel thresholdng, Adaptve PSO and basc PSO for House and Elane mages Images House Elane c Proposed thresholds 97,6 7,9,7 70,7,,89 6,0,6,67,96 08,67 96,,86 8,,,96 8,,,7,0 Proposed Unformty values Adaptve PSO based Basc PSO June 0 IJST, Transactons of Electrcal Engneerng, Volume 6, Number E

10 6 S. A. Mohammad et al. (a) (b) (c) (d) Fg.. Gray scale mage and hstogram of a) Orgnal Lena b) Proposed Method c) Adaptve PSO d) GA for c= thresholds (a) (b) (c) (d) Fg.. Gray scale mage of and hstogram a) Orgnal Peppers b) Proposed c) Adaptve PSO d) GA for c= thresholds The hgher value of unformty measure means that the qualty of found thresholds by that algorthm s superor to others. In other words, t s a knd of class separablty crteron whch uses wthn-class and between-class varances for ts formulaton. If the wthn-class varance s mnmal and that of the between-class s maxmal then the class separaton s satsfactory. Therefore, the proposed algorthm represents a good separablty accordng to ts unformty values whch means that t s superor to other algorthms. IJST, Transactons of Electrcal Engneerng, Volume 6, Number E June 0

11 An effcent based on ABC for 7 (a) (b) (c) Fg.. Gray scale mage and hstogram of a) Orgnal House b) Proposed Method c) Adaptve PSO for c= thresholds 6. CONCLUSION In ths study, we proposed a new optmal multlevel thresholdng based on ABC algorthm and Otsu s. An teratve ntalzng was also used to make the algorthm converge faster. Ths dea makes the computaton complexty grow lnearly when hgher number of thresholds s selected. Owng to ths property, ths algorthm could be used n real stuatons. The results show that ths s superor to other well-known and popular s proposed prevously. (a) (b) (c) Fg. 6. Gray scale mage and hstogram of a) Orgnal Elane b) Proposed Method c) Adaptve PSO for c= thresholds June 0 IJST, Transactons of Electrcal Engneerng, Volume 6, Number E

12 8 S. A. Mohammad et al. REFERENCES. Sezgn, M. & Sankur, B. (00), Survey over mage thresholdng technques and quanttatve performance evaluaton. Journal of Electronc Imagng, Vol., pp Sahoo, P. K., Soltan, S., Wong, A. K. C. & Chen, Y. C. (988). A survey of thresholdng technques, Computer Vson Graphcs Image Process, Vol., pp Lee, S. U., Chung, S. Y. & Park, R. H. (990). A comparatve performance study of several global thresholdng technques for segmentaton, Computer Vson Graphcs Image Process, Vol., pp Kapur, J. N., Sahoo, P. K. & Wong, A. K. C. (98). A new for gray-level pcture thresholdng usng the entropy of the hstogram. Computer Vson Graphcs Image Processng, Vol., pp Kttler, J. & Illngworth, J. (986). Mnmum error thresholdng. Pattern Recognton, Vol. 9, pp Pun, T. (98). Entropy thresholdng: A new approach. Computer Vson Graphcs Image Processng, Vol. 6, pp Pun, T. (980). A new for grey-level pcture thresholdng usng the entropy of the hstogram. Sgnal Processng, Vol., pp Otsu, N. (979). A threshold selecton from gray-level hstograms. IEEE Transactons on Systems, Man, Cybernet, SMC-9, Yn, P. Y. & Chen, L. H. (99). New for multlevel thresholdng usng the symmetry and dualty of the hstogram. Journal of Electroncs and Imagng, Vol., pp Tsa, D. M. (99). A fast thresholdng selecton procedure for multmodal and unmodal hstograms. Pattern Recognton Letters, Vol. 6, pp Lm, Y. K. & Lee, S. U. (990). On the color mage segmentaton algorthm based on the thresholdng and the fuzzy c-means technques. Pattern Recognton, Vol., pp Yn, P. Y. & Chen, L. H. (997). A fast teratve scheme for mult-level thresholdng s. Sgnal Processng 60, pp Karaboga, D. (00). An dea based on honey bee swarm for numercal optmzaton. Techncal Report-TR06, Ercyes Unversty, Engneerng Faculty, Computer Engneerng Department.. Chander, A., Chatterjee, A. & Sarry, P. (0). A new socal and momentum component adaptve PSO algorthm for mage segmentaton. Expert Systems wth Applcatons, Vol. 8, Issue, pp Akbar, R., Mohammad, A. & Zarat, K. (00). A novel bee swarm optmzaton algorthm for numercal functon optmzaton. Communcatons n Nonlnear Scence and Numercal Smulaton, Vol., pp.. 6. Hatam, M. & Masnad-Sraz, M. A. (008). Analytcal dscrete optmzaton. Iranan Journal of Scence & Technology, Transacton B, Engneerng, Vol., No. B, pp Akbar, R. & Zarat, K. (0). A Cooperatve approach to bee swarm optmzaton algorthm. Journal of Informaton Scence and Engneerng, Vol. 7, No., pp Nknam, T. & Golestaneh, F. (0). Enhanced bee swarm optmzaton algorthm for dynamc economc dspatch. IEEE Systems Journal, DOI: 0.09/JSYST.0.98, Issue Levne, M. D. & Nazf, A. M. (98). Dynamc measurement of computer generated mage segmentatons. IEEE Trans. Pattern Analyss and Machne Intellgence, Vol. 7, pp Ng, W. S. & Lee, C. K. (996). Comment on usng the unformty measure for performance measure n mage segmentaton. IEEE Transactons on Pattern Analyss And Machne Intellgence, Vol. 8, No. 9.. Karaboga, D. (00). An dea based on honey bee swarm for numercal optmzaton. Techncal Report-TR06, Ercyes Unversty, Engneerng Faculty, Computer Engneerng Department.. Basturk, B. & Karaboga, D. (006). An artfcal bee colony (ABC) algorthm for numerc functon optmzaton. IEEE Swarm Intellgence Symposum, Indana, USA.. Zarat, K., Akbar, R. & Zegham, V. (0). On the performance of bee algorthms for resource constraned project schedulng problem. Journal of Appled Soft Computng, Vol., No., pp IJST, Transactons of Electrcal Engneerng, Volume 6, Number E June 0

13 An effcent based on ABC for 9. Karaboga, D. & Basturk, B. (007). A powerful and effcent algorthm for numercal functon optmzaton: Artfcal Bee Colony (ABC) algorthm. J. Global Optm. Vol. 9, Issue, pp Akbar, R., Zegham, V. & Zarat, K. (0). Artfcal bee colony for resource constraned project schedulng problem. Internatonal Journal of Industral Engneerng Computatons, Vol., No., pp Karaboga, D. & Ozturk, C. (0). A novel clusterng approach: Artfcal Bee Colony (ABC) algorthm. Appled Soft Computng, Vol., pp Karaboga, D., Basturk, B. & Ozturk, C. (007). Artfcal Bee Colony (ABC) optmzaton algorthm for tranng feedforward neural networks, LNCS: Modelng Decsons for Artfcal Intellgence, MDAI. Vol. 67, Sprnger Verlag, pp Karaboga, D. & Ozturk, C. (009). Neural networks tranng by Artfcal Bee Colony Algorthm on pattern classfcaton. Neural Network World, Vol. 9, Issue, pp De Falco, I., Della Coppa, A. & Tarantno, E. (007). Facng classfcaton problems wth partcle swarm optmzaton. Journal of Appled Soft Computng, Vol. 7, Issue, pp Yn, P. Y. (999). A fast scheme for optmal thresholdng usng genetc algorthms. Sgnal Processng, Vol. 7, pp Yn, P. Y. & Chen, L. H. (99). New for multlevel thresholdng usng the symmetry and dualty of the hstogram. Journal of Electroncs and Imagng, Vol., pp ftp://sp.usc.edu/pub/database/msc.zp June 0 IJST, Transactons of Electrcal Engneerng, Volume 6, Number E

The Study of Teaching-learning-based Optimization Algorithm

The Study of Teaching-learning-based Optimization Algorithm Advanced Scence and Technology Letters Vol. (AST 06), pp.05- http://dx.do.org/0.57/astl.06. The Study of Teachng-learnng-based Optmzaton Algorthm u Sun, Yan fu, Lele Kong, Haolang Q,, Helongang Insttute

More information

Lecture Notes on Linear Regression

Lecture Notes on Linear Regression Lecture Notes on Lnear Regresson Feng L fl@sdueducn Shandong Unversty, Chna Lnear Regresson Problem In regresson problem, we am at predct a contnuous target value gven an nput feature vector We assume

More information

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur Module 3 LOSSY IMAGE COMPRESSION SYSTEMS Verson ECE IIT, Kharagpur Lesson 6 Theory of Quantzaton Verson ECE IIT, Kharagpur Instructonal Objectves At the end of ths lesson, the students should be able to:

More information

VQ widely used in coding speech, image, and video

VQ widely used in coding speech, image, and video at Scalar quantzers are specal cases of vector quantzers (VQ): they are constraned to look at one sample at a tme (memoryless) VQ does not have such constrant better RD perfomance expected Source codng

More information

Design and Optimization of Fuzzy Controller for Inverse Pendulum System Using Genetic Algorithm

Design and Optimization of Fuzzy Controller for Inverse Pendulum System Using Genetic Algorithm Desgn and Optmzaton of Fuzzy Controller for Inverse Pendulum System Usng Genetc Algorthm H. Mehraban A. Ashoor Unversty of Tehran Unversty of Tehran h.mehraban@ece.ut.ac.r a.ashoor@ece.ut.ac.r Abstract:

More information

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS Avalable onlne at http://sck.org J. Math. Comput. Sc. 3 (3), No., 6-3 ISSN: 97-537 COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

More information

A Novel Hybrid Bat Algorithm for the Multilevel Thresholding Medical Image Segmentation

A Novel Hybrid Bat Algorithm for the Multilevel Thresholding Medical Image Segmentation RESEARCH ARTICLE Copyrght 2015 Amercan Scentfc Publshers All rghts reserved Prnted n the Unted States of Amerca Journal of Medcal Imagng and Health Informatcs Vol. 5, 1 5, 2015 A Novel Hybrd Bat Algorthm

More information

Particle Swarm Optimization with Adaptive Mutation in Local Best of Particles

Particle Swarm Optimization with Adaptive Mutation in Local Best of Particles 1 Internatonal Congress on Informatcs, Envronment, Energy and Applcatons-IEEA 1 IPCSIT vol.38 (1) (1) IACSIT Press, Sngapore Partcle Swarm Optmzaton wth Adaptve Mutaton n Local Best of Partcles Nanda ulal

More information

CONTRAST ENHANCEMENT FOR MIMIMUM MEAN BRIGHTNESS ERROR FROM HISTOGRAM PARTITIONING INTRODUCTION

CONTRAST ENHANCEMENT FOR MIMIMUM MEAN BRIGHTNESS ERROR FROM HISTOGRAM PARTITIONING INTRODUCTION CONTRAST ENHANCEMENT FOR MIMIMUM MEAN BRIGHTNESS ERROR FROM HISTOGRAM PARTITIONING N. Phanthuna 1,2, F. Cheevasuvt 2 and S. Chtwong 2 1 Department of Electrcal Engneerng, Faculty of Engneerng Rajamangala

More information

Image Segmentation Using Thresholding and Swarm Intelligence

Image Segmentation Using Thresholding and Swarm Intelligence 74 JOURNA OF SOFTWARE, VO. 7, NO. 5, MAY Image Segmentaton Usng Thresholdng and Swarm Intellgence Zhwe Ye Hube Unversty of Technology / School of Computer Scence, Wuhan, Chna Emal: wezhye@63.com Zhengbng

More information

Kernel Methods and SVMs Extension

Kernel Methods and SVMs Extension Kernel Methods and SVMs Extenson The purpose of ths document s to revew materal covered n Machne Learnng 1 Supervsed Learnng regardng support vector machnes (SVMs). Ths document also provdes a general

More information

2E Pattern Recognition Solutions to Introduction to Pattern Recognition, Chapter 2: Bayesian pattern classification

2E Pattern Recognition Solutions to Introduction to Pattern Recognition, Chapter 2: Bayesian pattern classification E395 - Pattern Recognton Solutons to Introducton to Pattern Recognton, Chapter : Bayesan pattern classfcaton Preface Ths document s a soluton manual for selected exercses from Introducton to Pattern Recognton

More information

Lecture 10 Support Vector Machines II

Lecture 10 Support Vector Machines II Lecture 10 Support Vector Machnes II 22 February 2016 Taylor B. Arnold Yale Statstcs STAT 365/665 1/28 Notes: Problem 3 s posted and due ths upcomng Frday There was an early bug n the fake-test data; fxed

More information

CIS526: Machine Learning Lecture 3 (Sept 16, 2003) Linear Regression. Preparation help: Xiaoying Huang. x 1 θ 1 output... θ M x M

CIS526: Machine Learning Lecture 3 (Sept 16, 2003) Linear Regression. Preparation help: Xiaoying Huang. x 1 θ 1 output... θ M x M CIS56: achne Learnng Lecture 3 (Sept 6, 003) Preparaton help: Xaoyng Huang Lnear Regresson Lnear regresson can be represented by a functonal form: f(; θ) = θ 0 0 +θ + + θ = θ = 0 ote: 0 s a dummy attrbute

More information

Problem Set 9 Solutions

Problem Set 9 Solutions Desgn and Analyss of Algorthms May 4, 2015 Massachusetts Insttute of Technology 6.046J/18.410J Profs. Erk Demane, Srn Devadas, and Nancy Lynch Problem Set 9 Solutons Problem Set 9 Solutons Ths problem

More information

Chapter - 2. Distribution System Power Flow Analysis

Chapter - 2. Distribution System Power Flow Analysis Chapter - 2 Dstrbuton System Power Flow Analyss CHAPTER - 2 Radal Dstrbuton System Load Flow 2.1 Introducton Load flow s an mportant tool [66] for analyzng electrcal power system network performance. Load

More information

A PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS

A PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS HCMC Unversty of Pedagogy Thong Nguyen Huu et al. A PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS Thong Nguyen Huu and Hao Tran Van Department of mathematcs-nformaton,

More information

ECE559VV Project Report

ECE559VV Project Report ECE559VV Project Report (Supplementary Notes Loc Xuan Bu I. MAX SUM-RATE SCHEDULING: THE UPLINK CASE We have seen (n the presentaton that, for downlnk (broadcast channels, the strategy maxmzng the sum-rate

More information

For now, let us focus on a specific model of neurons. These are simplified from reality but can achieve remarkable results.

For now, let us focus on a specific model of neurons. These are simplified from reality but can achieve remarkable results. Neural Networks : Dervaton compled by Alvn Wan from Professor Jtendra Malk s lecture Ths type of computaton s called deep learnng and s the most popular method for many problems, such as computer vson

More information

Pop-Click Noise Detection Using Inter-Frame Correlation for Improved Portable Auditory Sensing

Pop-Click Noise Detection Using Inter-Frame Correlation for Improved Portable Auditory Sensing Advanced Scence and Technology Letters, pp.164-168 http://dx.do.org/10.14257/astl.2013 Pop-Clc Nose Detecton Usng Inter-Frame Correlaton for Improved Portable Audtory Sensng Dong Yun Lee, Kwang Myung Jeon,

More information

On the Multicriteria Integer Network Flow Problem

On the Multicriteria Integer Network Flow Problem BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 5, No 2 Sofa 2005 On the Multcrtera Integer Network Flow Problem Vassl Vasslev, Marana Nkolova, Maryana Vassleva Insttute of

More information

Regularized Discriminant Analysis for Face Recognition

Regularized Discriminant Analysis for Face Recognition 1 Regularzed Dscrmnant Analyss for Face Recognton Itz Pma, Mayer Aladem Department of Electrcal and Computer Engneerng, Ben-Guron Unversty of the Negev P.O.Box 653, Beer-Sheva, 845, Israel. Abstract Ths

More information

Assortment Optimization under MNL

Assortment Optimization under MNL Assortment Optmzaton under MNL Haotan Song Aprl 30, 2017 1 Introducton The assortment optmzaton problem ams to fnd the revenue-maxmzng assortment of products to offer when the prces of products are fxed.

More information

Appendix B: Resampling Algorithms

Appendix B: Resampling Algorithms 407 Appendx B: Resamplng Algorthms A common problem of all partcle flters s the degeneracy of weghts, whch conssts of the unbounded ncrease of the varance of the mportance weghts ω [ ] of the partcles

More information

Generalized Linear Methods

Generalized Linear Methods Generalzed Lnear Methods 1 Introducton In the Ensemble Methods the general dea s that usng a combnaton of several weak learner one could make a better learner. More formally, assume that we have a set

More information

Boostrapaggregating (Bagging)

Boostrapaggregating (Bagging) Boostrapaggregatng (Baggng) An ensemble meta-algorthm desgned to mprove the stablty and accuracy of machne learnng algorthms Can be used n both regresson and classfcaton Reduces varance and helps to avod

More information

A Bayes Algorithm for the Multitask Pattern Recognition Problem Direct Approach

A Bayes Algorithm for the Multitask Pattern Recognition Problem Direct Approach A Bayes Algorthm for the Multtask Pattern Recognton Problem Drect Approach Edward Puchala Wroclaw Unversty of Technology, Char of Systems and Computer etworks, Wybrzeze Wyspanskego 7, 50-370 Wroclaw, Poland

More information

A Hybrid Variational Iteration Method for Blasius Equation

A Hybrid Variational Iteration Method for Blasius Equation Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 1 (June 2015), pp. 223-229 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) A Hybrd Varatonal Iteraton Method

More information

Logistic Regression. CAP 5610: Machine Learning Instructor: Guo-Jun QI

Logistic Regression. CAP 5610: Machine Learning Instructor: Guo-Jun QI Logstc Regresson CAP 561: achne Learnng Instructor: Guo-Jun QI Bayes Classfer: A Generatve model odel the posteror dstrbuton P(Y X) Estmate class-condtonal dstrbuton P(X Y) for each Y Estmate pror dstrbuton

More information

Multigradient for Neural Networks for Equalizers 1

Multigradient for Neural Networks for Equalizers 1 Multgradent for Neural Netorks for Equalzers 1 Chulhee ee, Jnook Go and Heeyoung Km Department of Electrcal and Electronc Engneerng Yonse Unversty 134 Shnchon-Dong, Seodaemun-Ku, Seoul 1-749, Korea ABSTRACT

More information

The Synchronous 8th-Order Differential Attack on 12 Rounds of the Block Cipher HyRAL

The Synchronous 8th-Order Differential Attack on 12 Rounds of the Block Cipher HyRAL The Synchronous 8th-Order Dfferental Attack on 12 Rounds of the Block Cpher HyRAL Yasutaka Igarash, Sej Fukushma, and Tomohro Hachno Kagoshma Unversty, Kagoshma, Japan Emal: {garash, fukushma, hachno}@eee.kagoshma-u.ac.jp

More information

Solving Nonlinear Differential Equations by a Neural Network Method

Solving Nonlinear Differential Equations by a Neural Network Method Solvng Nonlnear Dfferental Equatons by a Neural Network Method Luce P. Aarts and Peter Van der Veer Delft Unversty of Technology, Faculty of Cvlengneerng and Geoscences, Secton of Cvlengneerng Informatcs,

More information

The Minimum Universal Cost Flow in an Infeasible Flow Network

The Minimum Universal Cost Flow in an Infeasible Flow Network Journal of Scences, Islamc Republc of Iran 17(2): 175-180 (2006) Unversty of Tehran, ISSN 1016-1104 http://jscencesutacr The Mnmum Unversal Cost Flow n an Infeasble Flow Network H Saleh Fathabad * M Bagheran

More information

Chapter Newton s Method

Chapter Newton s Method Chapter 9. Newton s Method After readng ths chapter, you should be able to:. Understand how Newton s method s dfferent from the Golden Secton Search method. Understand how Newton s method works 3. Solve

More information

A Hybrid Mutation-based Artificial Bee Colony for Traveling Salesman Problem

A Hybrid Mutation-based Artificial Bee Colony for Traveling Salesman Problem A Hybrd Mutaton-based Artfcal Bee Colony for Travelng Salesman Problem Shma Sabet and Fardad Farokh Islamc Azad Unversty of Central Tehran Branch, Tehran, Iran Emal: sh_sabet25@yahoo.com, f_farokh@auctb.ac.r

More information

Multilayer Perceptron (MLP)

Multilayer Perceptron (MLP) Multlayer Perceptron (MLP) Seungjn Cho Department of Computer Scence and Engneerng Pohang Unversty of Scence and Technology 77 Cheongam-ro, Nam-gu, Pohang 37673, Korea seungjn@postech.ac.kr 1 / 20 Outlne

More information

EEL 6266 Power System Operation and Control. Chapter 3 Economic Dispatch Using Dynamic Programming

EEL 6266 Power System Operation and Control. Chapter 3 Economic Dispatch Using Dynamic Programming EEL 6266 Power System Operaton and Control Chapter 3 Economc Dspatch Usng Dynamc Programmng Pecewse Lnear Cost Functons Common practce many utltes prefer to represent ther generator cost functons as sngle-

More information

Single-Facility Scheduling over Long Time Horizons by Logic-based Benders Decomposition

Single-Facility Scheduling over Long Time Horizons by Logic-based Benders Decomposition Sngle-Faclty Schedulng over Long Tme Horzons by Logc-based Benders Decomposton Elvn Coban and J. N. Hooker Tepper School of Busness, Carnege Mellon Unversty ecoban@andrew.cmu.edu, john@hooker.tepper.cmu.edu

More information

Solving of Single-objective Problems based on a Modified Multiple-crossover Genetic Algorithm: Test Function Study

Solving of Single-objective Problems based on a Modified Multiple-crossover Genetic Algorithm: Test Function Study Internatonal Conference on Systems, Sgnal Processng and Electroncs Engneerng (ICSSEE'0 December 6-7, 0 Duba (UAE Solvng of Sngle-objectve Problems based on a Modfed Multple-crossover Genetc Algorthm: Test

More information

Performance Analysis of Data Clustering Algorithms using Various Effectiveness Measures

Performance Analysis of Data Clustering Algorithms using Various Effectiveness Measures 336 The Internatonal Arab Journal of Informaton Technology VOL. 3, NO. 3, May 06 Performance Analyss of Data Clusterng Algorthms usng Varous Effectveness Measures Krshnamoorth Murugasamy and Natarajan

More information

EEE 241: Linear Systems

EEE 241: Linear Systems EEE : Lnear Systems Summary #: Backpropagaton BACKPROPAGATION The perceptron rule as well as the Wdrow Hoff learnng were desgned to tran sngle layer networks. They suffer from the same dsadvantage: they

More information

Entropy Generation Minimization of Pin Fin Heat Sinks by Means of Metaheuristic Methods

Entropy Generation Minimization of Pin Fin Heat Sinks by Means of Metaheuristic Methods Indan Journal of Scence and Technology Entropy Generaton Mnmzaton of Pn Fn Heat Snks by Means of Metaheurstc Methods Amr Jafary Moghaddam * and Syfollah Saedodn Department of Mechancal Engneerng, Semnan

More information

Using Immune Genetic Algorithm to Optimize BP Neural Network and Its Application Peng-fei LIU1,Qun-tai SHEN1 and Jun ZHI2,*

Using Immune Genetic Algorithm to Optimize BP Neural Network and Its Application Peng-fei LIU1,Qun-tai SHEN1 and Jun ZHI2,* Advances n Computer Scence Research (ACRS), volume 54 Internatonal Conference on Computer Networks and Communcaton Technology (CNCT206) Usng Immune Genetc Algorthm to Optmze BP Neural Network and Its Applcaton

More information

A Hybrid Differential Evolution Algorithm Game Theory for the Berth Allocation Problem

A Hybrid Differential Evolution Algorithm Game Theory for the Berth Allocation Problem A Hybrd Dfferental Evoluton Algorthm ame Theory for the Berth Allocaton Problem Nasser R. Sabar, Sang Yew Chong, and raham Kendall The Unversty of Nottngham Malaysa Campus, Jalan Broga, 43500 Semenyh,

More information

Resource Allocation with a Budget Constraint for Computing Independent Tasks in the Cloud

Resource Allocation with a Budget Constraint for Computing Independent Tasks in the Cloud Resource Allocaton wth a Budget Constrant for Computng Independent Tasks n the Cloud Wemng Sh and Bo Hong School of Electrcal and Computer Engneerng Georga Insttute of Technology, USA 2nd IEEE Internatonal

More information

Homework Assignment 3 Due in class, Thursday October 15

Homework Assignment 3 Due in class, Thursday October 15 Homework Assgnment 3 Due n class, Thursday October 15 SDS 383C Statstcal Modelng I 1 Rdge regresson and Lasso 1. Get the Prostrate cancer data from http://statweb.stanford.edu/~tbs/elemstatlearn/ datasets/prostate.data.

More information

A Fast Fractal Image Compression Algorithm Using Predefined Values for Contrast Scaling

A Fast Fractal Image Compression Algorithm Using Predefined Values for Contrast Scaling Proceedngs of the World Congress on Engneerng and Computer Scence 007 WCECS 007, October 4-6, 007, San Francsco, USA A Fast Fractal Image Compresson Algorthm Usng Predefned Values for Contrast Scalng H.

More information

FUZZY GOAL PROGRAMMING VS ORDINARY FUZZY PROGRAMMING APPROACH FOR MULTI OBJECTIVE PROGRAMMING PROBLEM

FUZZY GOAL PROGRAMMING VS ORDINARY FUZZY PROGRAMMING APPROACH FOR MULTI OBJECTIVE PROGRAMMING PROBLEM Internatonal Conference on Ceramcs, Bkaner, Inda Internatonal Journal of Modern Physcs: Conference Seres Vol. 22 (2013) 757 761 World Scentfc Publshng Company DOI: 10.1142/S2010194513010982 FUZZY GOAL

More information

A New Scrambling Evaluation Scheme based on Spatial Distribution Entropy and Centroid Difference of Bit-plane

A New Scrambling Evaluation Scheme based on Spatial Distribution Entropy and Centroid Difference of Bit-plane A New Scramblng Evaluaton Scheme based on Spatal Dstrbuton Entropy and Centrod Dfference of Bt-plane Lang Zhao *, Avshek Adhkar Kouch Sakura * * Graduate School of Informaton Scence and Electrcal Engneerng,

More information

Optimal multilevel thresholding based on molecular kinetic theory optimization algorithm and line intercept histogram

Optimal multilevel thresholding based on molecular kinetic theory optimization algorithm and line intercept histogram DOI:.7/s77-6-335-8 Optmal multlevel thresholdng based on molecular knetc theory optmzaton algorthm and lne ntercept hstogram FAN Chao-dong( 范朝冬 ), 2, REN Ke( 任柯 ), 2, ZHANG Yng-je( 张英杰 ) 3, YI Lng-zh(

More information

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin Proceedngs of the 007 Wnter Smulaton Conference S G Henderson, B Bller, M-H Hseh, J Shortle, J D Tew, and R R Barton, eds LOW BIAS INTEGRATED PATH ESTIMATORS James M Calvn Department of Computer Scence

More information

Interactive Bi-Level Multi-Objective Integer. Non-linear Programming Problem

Interactive Bi-Level Multi-Objective Integer. Non-linear Programming Problem Appled Mathematcal Scences Vol 5 0 no 65 3 33 Interactve B-Level Mult-Objectve Integer Non-lnear Programmng Problem O E Emam Department of Informaton Systems aculty of Computer Scence and nformaton Helwan

More information

Comparison of the Population Variance Estimators. of 2-Parameter Exponential Distribution Based on. Multiple Criteria Decision Making Method

Comparison of the Population Variance Estimators. of 2-Parameter Exponential Distribution Based on. Multiple Criteria Decision Making Method Appled Mathematcal Scences, Vol. 7, 0, no. 47, 07-0 HIARI Ltd, www.m-hkar.com Comparson of the Populaton Varance Estmators of -Parameter Exponental Dstrbuton Based on Multple Crtera Decson Makng Method

More information

MULTISPECTRAL IMAGE CLASSIFICATION USING BACK-PROPAGATION NEURAL NETWORK IN PCA DOMAIN

MULTISPECTRAL IMAGE CLASSIFICATION USING BACK-PROPAGATION NEURAL NETWORK IN PCA DOMAIN MULTISPECTRAL IMAGE CLASSIFICATION USING BACK-PROPAGATION NEURAL NETWORK IN PCA DOMAIN S. Chtwong, S. Wtthayapradt, S. Intajag, and F. Cheevasuvt Faculty of Engneerng, Kng Mongkut s Insttute of Technology

More information

Lecture 20: November 7

Lecture 20: November 7 0-725/36-725: Convex Optmzaton Fall 205 Lecturer: Ryan Tbshran Lecture 20: November 7 Scrbes: Varsha Chnnaobreddy, Joon Sk Km, Lngyao Zhang Note: LaTeX template courtesy of UC Berkeley EECS dept. Dsclamer:

More information

Lecture 12: Classification

Lecture 12: Classification Lecture : Classfcaton g Dscrmnant functons g The optmal Bayes classfer g Quadratc classfers g Eucldean and Mahalanobs metrcs g K Nearest Neghbor Classfers Intellgent Sensor Systems Rcardo Guterrez-Osuna

More information

On the Interval Zoro Symmetric Single-step Procedure for Simultaneous Finding of Polynomial Zeros

On the Interval Zoro Symmetric Single-step Procedure for Simultaneous Finding of Polynomial Zeros Appled Mathematcal Scences, Vol. 5, 2011, no. 75, 3693-3706 On the Interval Zoro Symmetrc Sngle-step Procedure for Smultaneous Fndng of Polynomal Zeros S. F. M. Rusl, M. Mons, M. A. Hassan and W. J. Leong

More information

Discretization of Continuous Attributes in Rough Set Theory and Its Application*

Discretization of Continuous Attributes in Rough Set Theory and Its Application* Dscretzaton of Contnuous Attrbutes n Rough Set Theory and Its Applcaton* Gexang Zhang 1,2, Lazhao Hu 1, and Wedong Jn 2 1 Natonal EW Laboratory, Chengdu 610036 Schuan, Chna dylan7237@sna.com 2 School of

More information

Lossy Compression. Compromise accuracy of reconstruction for increased compression.

Lossy Compression. Compromise accuracy of reconstruction for increased compression. Lossy Compresson Compromse accuracy of reconstructon for ncreased compresson. The reconstructon s usually vsbly ndstngushable from the orgnal mage. Typcally, one can get up to 0:1 compresson wth almost

More information

Multilevel Thresholding for Image Segmentation using the Galaxy-based Search Algorithm

Multilevel Thresholding for Image Segmentation using the Galaxy-based Search Algorithm I.J. Intellgent Systems and Applcatons, 2013, 11, 19-33 Publshed Onlne October 2013 n MECS (http://www.mecs-press.org/) DOI: 10.5815/jsa.2013.11.03 Multlevel Thresholdng for Segmentaton usng the Galaxy-based

More information

Markov Chain Monte Carlo Lecture 6

Markov Chain Monte Carlo Lecture 6 where (x 1,..., x N ) X N, N s called the populaton sze, f(x) f (x) for at least one {1, 2,..., N}, and those dfferent from f(x) are called the tral dstrbutons n terms of mportance samplng. Dfferent ways

More information

Natural Language Processing and Information Retrieval

Natural Language Processing and Information Retrieval Natural Language Processng and Informaton Retreval Support Vector Machnes Alessandro Moschtt Department of nformaton and communcaton technology Unversty of Trento Emal: moschtt@ds.untn.t Summary Support

More information

An Admission Control Algorithm in Cloud Computing Systems

An Admission Control Algorithm in Cloud Computing Systems An Admsson Control Algorthm n Cloud Computng Systems Authors: Frank Yeong-Sung Ln Department of Informaton Management Natonal Tawan Unversty Tape, Tawan, R.O.C. ysln@m.ntu.edu.tw Yngje Lan Management Scence

More information

The Expectation-Maximization Algorithm

The Expectation-Maximization Algorithm The Expectaton-Maxmaton Algorthm Charles Elan elan@cs.ucsd.edu November 16, 2007 Ths chapter explans the EM algorthm at multple levels of generalty. Secton 1 gves the standard hgh-level verson of the algorthm.

More information

Real-Time Systems. Multiprocessor scheduling. Multiprocessor scheduling. Multiprocessor scheduling

Real-Time Systems. Multiprocessor scheduling. Multiprocessor scheduling. Multiprocessor scheduling Real-Tme Systems Multprocessor schedulng Specfcaton Implementaton Verfcaton Multprocessor schedulng -- -- Global schedulng How are tasks assgned to processors? Statc assgnment The processor(s) used for

More information

Finite Mixture Models and Expectation Maximization. Most slides are from: Dr. Mario Figueiredo, Dr. Anil Jain and Dr. Rong Jin

Finite Mixture Models and Expectation Maximization. Most slides are from: Dr. Mario Figueiredo, Dr. Anil Jain and Dr. Rong Jin Fnte Mxture Models and Expectaton Maxmzaton Most sldes are from: Dr. Maro Fgueredo, Dr. Anl Jan and Dr. Rong Jn Recall: The Supervsed Learnng Problem Gven a set of n samples X {(x, y )},,,n Chapter 3 of

More information

Annexes. EC.1. Cycle-base move illustration. EC.2. Problem Instances

Annexes. EC.1. Cycle-base move illustration. EC.2. Problem Instances ec Annexes Ths Annex frst llustrates a cycle-based move n the dynamc-block generaton tabu search. It then dsplays the characterstcs of the nstance sets, followed by detaled results of the parametercalbraton

More information

Support Vector Machines. Vibhav Gogate The University of Texas at dallas

Support Vector Machines. Vibhav Gogate The University of Texas at dallas Support Vector Machnes Vbhav Gogate he Unversty of exas at dallas What We have Learned So Far? 1. Decson rees. Naïve Bayes 3. Lnear Regresson 4. Logstc Regresson 5. Perceptron 6. Neural networks 7. K-Nearest

More information

An Adaptive Learning Particle Swarm Optimizer for Function Optimization

An Adaptive Learning Particle Swarm Optimizer for Function Optimization An Adaptve Learnng Partcle Swarm Optmzer for Functon Optmzaton Changhe L and Shengxang Yang Abstract Tradtonal partcle swarm optmzaton (PSO) suffers from the premature convergence problem, whch usually

More information

Expectation Maximization Mixture Models HMMs

Expectation Maximization Mixture Models HMMs -755 Machne Learnng for Sgnal Processng Mture Models HMMs Class 9. 2 Sep 200 Learnng Dstrbutons for Data Problem: Gven a collecton of eamples from some data, estmate ts dstrbuton Basc deas of Mamum Lelhood

More information

CSC 411 / CSC D11 / CSC C11

CSC 411 / CSC D11 / CSC C11 18 Boostng s a general strategy for learnng classfers by combnng smpler ones. The dea of boostng s to take a weak classfer that s, any classfer that wll do at least slghtly better than chance and use t

More information

Polynomial Regression Models

Polynomial Regression Models LINEAR REGRESSION ANALYSIS MODULE XII Lecture - 6 Polynomal Regresson Models Dr. Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur Test of sgnfcance To test the sgnfcance

More information

An Improved multiple fractal algorithm

An Improved multiple fractal algorithm Advanced Scence and Technology Letters Vol.31 (MulGraB 213), pp.184-188 http://dx.do.org/1.1427/astl.213.31.41 An Improved multple fractal algorthm Yun Ln, Xaochu Xu, Jnfeng Pang College of Informaton

More information

COS 521: Advanced Algorithms Game Theory and Linear Programming

COS 521: Advanced Algorithms Game Theory and Linear Programming COS 521: Advanced Algorthms Game Theory and Lnear Programmng Moses Charkar February 27, 2013 In these notes, we ntroduce some basc concepts n game theory and lnear programmng (LP). We show a connecton

More information

Pattern Classification

Pattern Classification Pattern Classfcaton All materals n these sldes ere taken from Pattern Classfcaton (nd ed) by R. O. Duda, P. E. Hart and D. G. Stork, John Wley & Sons, 000 th the permsson of the authors and the publsher

More information

The Convergence Speed of Single- And Multi-Objective Immune Algorithm Based Optimization Problems

The Convergence Speed of Single- And Multi-Objective Immune Algorithm Based Optimization Problems The Convergence Speed of Sngle- And Mult-Obectve Immune Algorthm Based Optmzaton Problems Mohammed Abo-Zahhad Faculty of Engneerng, Electrcal and Electroncs Engneerng Department, Assut Unversty, Assut,

More information

Research on Route guidance of logistic scheduling problem under fuzzy time window

Research on Route guidance of logistic scheduling problem under fuzzy time window Advanced Scence and Technology Letters, pp.21-30 http://dx.do.org/10.14257/astl.2014.78.05 Research on Route gudance of logstc schedulng problem under fuzzy tme wndow Yuqang Chen 1, Janlan Guo 2 * Department

More information

Ensemble Methods: Boosting

Ensemble Methods: Boosting Ensemble Methods: Boostng Ncholas Ruozz Unversty of Texas at Dallas Based on the sldes of Vbhav Gogate and Rob Schapre Last Tme Varance reducton va baggng Generate new tranng data sets by samplng wth replacement

More information

Tracking with Kalman Filter

Tracking with Kalman Filter Trackng wth Kalman Flter Scott T. Acton Vrgna Image and Vdeo Analyss (VIVA), Charles L. Brown Department of Electrcal and Computer Engneerng Department of Bomedcal Engneerng Unversty of Vrgna, Charlottesvlle,

More information

Cluster Validation Determining Number of Clusters. Umut ORHAN, PhD.

Cluster Validation Determining Number of Clusters. Umut ORHAN, PhD. Cluster Analyss Cluster Valdaton Determnng Number of Clusters 1 Cluster Valdaton The procedure of evaluatng the results of a clusterng algorthm s known under the term cluster valdty. How do we evaluate

More information

A New Evolutionary Computation Based Approach for Learning Bayesian Network

A New Evolutionary Computation Based Approach for Learning Bayesian Network Avalable onlne at www.scencedrect.com Proceda Engneerng 15 (2011) 4026 4030 Advanced n Control Engneerng and Informaton Scence A New Evolutonary Computaton Based Approach for Learnng Bayesan Network Yungang

More information

Winter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan

Winter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan Wnter 2008 CS567 Stochastc Lnear/Integer Programmng Guest Lecturer: Xu, Huan Class 2: More Modelng Examples 1 Capacty Expanson Capacty expanson models optmal choces of the tmng and levels of nvestments

More information

Probability Density Function Estimation by different Methods

Probability Density Function Estimation by different Methods EEE 739Q SPRIG 00 COURSE ASSIGMET REPORT Probablty Densty Functon Estmaton by dfferent Methods Vas Chandraant Rayar Abstract The am of the assgnment was to estmate the probablty densty functon (PDF of

More information

A linear imaging system with white additive Gaussian noise on the observed data is modeled as follows:

A linear imaging system with white additive Gaussian noise on the observed data is modeled as follows: Supplementary Note Mathematcal bacground A lnear magng system wth whte addtve Gaussan nose on the observed data s modeled as follows: X = R ϕ V + G, () where X R are the expermental, two-dmensonal proecton

More information

Segmentation of Infrared Images and Objectives Detection Using Maximum Entropy Method Based on the Bee Algorithm

Segmentation of Infrared Images and Objectives Detection Using Maximum Entropy Method Based on the Bee Algorithm Internatonal Journal of Computer Informaton Systems and Industral Management Applcatons (IJCISIM) ISSN: 2150-7988 Vol.3 (2011), pp.026-033 http://www.mrlabs.org/jcsm Segmentaton of Infrared Images and

More information

Riccardo Poli, James Kennedy, Tim Blackwell: Particle swarm optimization. Swarm Intelligence 1(1): (2007)

Riccardo Poli, James Kennedy, Tim Blackwell: Particle swarm optimization. Swarm Intelligence 1(1): (2007) Sldes largely based on: Rccardo Pol, James Kennedy, Tm Blackwell: Partcle swarm optmzaton. Swarm Intellgence 1(1): 33-57 (2007) Partcle Swarm Optmzaton Sldes largely based on: Rccardo Pol, James Kennedy,

More information

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009 College of Computer & Informaton Scence Fall 2009 Northeastern Unversty 20 October 2009 CS7880: Algorthmc Power Tools Scrbe: Jan Wen and Laura Poplawsk Lecture Outlne: Prmal-dual schema Network Desgn:

More information

CHAPTER 7 STOCHASTIC ECONOMIC EMISSION DISPATCH-MODELED USING WEIGHTING METHOD

CHAPTER 7 STOCHASTIC ECONOMIC EMISSION DISPATCH-MODELED USING WEIGHTING METHOD 90 CHAPTER 7 STOCHASTIC ECOOMIC EMISSIO DISPATCH-MODELED USIG WEIGHTIG METHOD 7.1 ITRODUCTIO early 70% of electrc power produced n the world s by means of thermal plants. Thermal power statons are the

More information

Optimal Placement and Sizing of DGs in the Distribution System for Loss Minimization and Voltage Stability Improvement using CABC

Optimal Placement and Sizing of DGs in the Distribution System for Loss Minimization and Voltage Stability Improvement using CABC Internatonal Journal on Electrcal Engneerng and Informatcs - Volume 7, Number 4, Desember 2015 Optmal Placement and Szng of s n the Dstrbuton System for Loss Mnmzaton and Voltage Stablty Improvement usng

More information

An Upper Bound on SINR Threshold for Call Admission Control in Multiple-Class CDMA Systems with Imperfect Power-Control

An Upper Bound on SINR Threshold for Call Admission Control in Multiple-Class CDMA Systems with Imperfect Power-Control An Upper Bound on SINR Threshold for Call Admsson Control n Multple-Class CDMA Systems wth Imperfect ower-control Mahmoud El-Sayes MacDonald, Dettwler and Assocates td. (MDA) Toronto, Canada melsayes@hotmal.com

More information

Some modelling aspects for the Matlab implementation of MMA

Some modelling aspects for the Matlab implementation of MMA Some modellng aspects for the Matlab mplementaton of MMA Krster Svanberg krlle@math.kth.se Optmzaton and Systems Theory Department of Mathematcs KTH, SE 10044 Stockholm September 2004 1. Consdered optmzaton

More information

Comparative Analysis of SPSO and PSO to Optimal Power Flow Solutions

Comparative Analysis of SPSO and PSO to Optimal Power Flow Solutions Internatonal Journal for Research n Appled Scence & Engneerng Technology (IJRASET) Volume 6 Issue I, January 018- Avalable at www.jraset.com Comparatve Analyss of SPSO and PSO to Optmal Power Flow Solutons

More information

Global Sensitivity. Tuesday 20 th February, 2018

Global Sensitivity. Tuesday 20 th February, 2018 Global Senstvty Tuesday 2 th February, 28 ) Local Senstvty Most senstvty analyses [] are based on local estmates of senstvty, typcally by expandng the response n a Taylor seres about some specfc values

More information

System identifications by SIRMs models with linear transformation of input variables

System identifications by SIRMs models with linear transformation of input variables ORIGINAL RESEARCH System dentfcatons by SIRMs models wth lnear transformaton of nput varables Hrofum Myama, Nortaka Shge, Hrom Myama Graduate School of Scence and Engneerng, Kagoshma Unversty, Japan Receved:

More information

Feature Selection: Part 1

Feature Selection: Part 1 CSE 546: Machne Learnng Lecture 5 Feature Selecton: Part 1 Instructor: Sham Kakade 1 Regresson n the hgh dmensonal settng How do we learn when the number of features d s greater than the sample sze n?

More information

Numerical Heat and Mass Transfer

Numerical Heat and Mass Transfer Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and

More information

10-701/ Machine Learning, Fall 2005 Homework 3

10-701/ Machine Learning, Fall 2005 Homework 3 10-701/15-781 Machne Learnng, Fall 2005 Homework 3 Out: 10/20/05 Due: begnnng of the class 11/01/05 Instructons Contact questons-10701@autonlaborg for queston Problem 1 Regresson and Cross-valdaton [40

More information

Which Separator? Spring 1

Which Separator? Spring 1 Whch Separator? 6.034 - Sprng 1 Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng 3 Margn of a pont " # y (w $ + b) proportonal

More information

IMAGE DENOISING USING NEW ADAPTIVE BASED MEDIAN FILTER

IMAGE DENOISING USING NEW ADAPTIVE BASED MEDIAN FILTER Sgnal & Image Processng : An Internatonal Journal (SIPIJ) Vol.5, No.4, August 2014 IMAGE DENOISING USING NEW ADAPTIVE BASED MEDIAN FILTER Suman Shrestha 1, 2 1 Unversty of Massachusetts Medcal School,

More information

Neural networks. Nuno Vasconcelos ECE Department, UCSD

Neural networks. Nuno Vasconcelos ECE Department, UCSD Neural networs Nuno Vasconcelos ECE Department, UCSD Classfcaton a classfcaton problem has two types of varables e.g. X - vector of observatons (features) n the world Y - state (class) of the world x X

More information