A general forulation of the cross-nested logit odel Michel Bierlaire, Dpt of Matheatics, EPFL, Lausanne Phone: Fax:

Size: px
Start display at page:

Download "A general forulation of the cross-nested logit odel Michel Bierlaire, Dpt of Matheatics, EPFL, Lausanne Phone: Fax:"

Transcription

1 A general forulation of the cross-nested logit odel Michel Bierlaire, EPFL Conference paper STRC 2001 Session: Choices

2 A general forulation of the cross-nested logit odel Michel Bierlaire, Dpt of Matheatics, EPFL, Lausanne Phone: Fax: Eail: Abstract The cross-nested logit (CNL) odel has been entioned for the first tie by Vovsha (1997) in the context of a ode choice survey in Israel. Actually, it is alost identical to the Ordered GEV odel proposed by Sall (1987). This odel, eber of the GEV faily (McFadden, 1978), is appealing for its ability to capture a wide variety of correlation structures. Papola (2000) has shown that a specific CNL odel can be derived for any given hooschedastic variance-covariance atrix. Therefore, the CNL odel, with a closed for forulation derived fro the GEV odel, becoes a serious copetitor for the probit odel. In this paper, we develop on the general forulation of the Cross Nested Logit odel proposed by Ben-Akiva and Bierlaire (1999). We show that the forulations by Sall (1987), Vovsha (1997) and Papola (2000) are particular cases of our general forulation. We also provide soe new insights in the CNL odel based on theoretical analysis and epirical tests. We show that one of the conditions iposed by Sall (1987), Vovsha (1997) and Papola (2000) is not necessary for the CNL odel to be consistent with rando utility theory. This condition iposes that the su of the paraeters capturing the degree of ebership of an alternative to a nest is equal to one. Weprove that the CNL odel is a GEV odel without using that condition. It is actually a noralization condition, iportant for paraeters identification, but not for odel forulation. Finally, we use a siplistic odel to illustrate the role of the noralization condition. We propose a variant of the condition proposed in the literature which is slightly ore general. Keywords: logit odel, cross-nested odel, GEV odel, rando utility, transportation deand Swiss Transport Research Conference STRC 2001 Monte Verita

3 1 Introduction Discrete choice odels play a aor role in any fields involving a huan diension, including transportation deand analysis (the recent Nobel prize to D. McFadden is a perfect illustration of that stateent). Their nice and strong theoretical properties, and their flexibility to capture various situations, provide a vast topic of interest for both researchers and practitioners, that has (by far) not been totally exploited yet. The particular structure of transportation related choice situations are not always fully consistent with the underlying odeling theory (Ben-Akiva and Bierlaire, 1999), requiring to enhance and adapt existing odels. The theory on GEV odels have been introduced by McFadden (1978). It provides a treendous potential, as it defines a whole faily of odels, consistent with rando utility theory. It appears that only a few ebers of this faily have been exploited so far. Aong the, the cross-nested logit (CNL) odel has been used by Vovsha (1997) in the context of a ode choice survey in Israel. Actually, Vovsha's odel is alost identical to the Ordered GEV odel proposed by Sall (1987). This odel is appealing for its ability to capture a wide variety of correlation structures. Papola (2000) has shown that a specific CNL odel can be obtained for any given hooschedastic variance-covariance atrix. Therefore, the CNL odel, with a closed for forulation derived fro the GEV odel, becoes a serious copetitor for the probit odel. It has been shown to be specifically appropriate for route choice applications (Vovsha and Bekhor, 1998), where topological correlations cannot be captured correctly by the ultinoial and the nested logit odels. Unfortunately, no satisfactory procedure has been proposed for the estiation of the odel paraeters. 1

4 The reaining of this section introduces the GEV odel and presents various forulations of the Cross-Nested Logit odel. In Section 2, we analyze the ost general forulation. We prove that it is consistent with the GEV odel faily. 1.1 The GEV odel The Generalized Extree Value (GEV) odel has been derived fro the rando utility odel by McFadden (1978). This general odel consists of a large faily of odels that include the Multinoial Logit and the Nested Logit odels. The probability of choosing alternative i within the choice set C of a given choice aker is P (ic) = i (y 1 ;:::;y J ) (1) G(y 1 ;:::;y J ) where J is the nuber of available alternatives, y i = e V i, V i is the deterinistic part of the utility function associated to alternative i, and G is a non-negative differentiable function defined on R J + with the following properties: 1. G is hoogeneous of degree >0, that is G(ffy) =ff G(y), 2. li yi!+1 G(y 1 ;:::;y i ;:::;y J )=+1, for each i =1;:::;J, 3. the kth partial derivative with respect to k distinct y i is non-negative ifk is odd and non-positive ifk is even that is, for any distinct indices i 1 ;:::;i k 2 f1;:::;jg, we have ( 1) k i1 :::@x ik (x)» 0; 8x 2 R J +: (2) Note that the hoogeneity of G and Euler's theore give P (ic) = e V i+ln G i (:::) P J =1 ev +ln G (:::) ; (3) where G i. It is well known that the Multinoial Logit and the Nested Logit odels are instances of this odel faily. We present now several forulations of the Cross-Nested Logit odel, derived fro the GEV odel. 2

5 1.2 Forulations of the Cross-Nested Logit odel The liitations of the Nested Logit odel has been observed by several authors (Willias, 1977, Forinash and Koppelan, 1993). The requireent ofunabiguous assignent of alternatives to nests does not allow to capture ixed interactions across alternatives. It sees that the first Cross-Nested Logit odel has been proposed by Sall (1987) in the context of departure tie choice. Sall's odel, called the Ordered GEV odel, isbased on the following function:! J+M ρr G(y 1 ;:::;y )= w r y 1=ρr ; (4) r=1 ψ Br where M is a positive integer, ρ r and w are constants satisfying 0 < ρ r» 1, w 0and M =0 The B r are overlapping subsets of alternatives: w =1: (5) B r = f 2f1;:::;Jgr M»» rg: (6) Vovsha (1997) applies the Cross-Nested Logit odel to a ode choice application, where the park&ride" alternative is allowed to belong to the coposite auto" and the coposite transit" nests. Vovsha derives the Cross-Nested Logit fro the GEV odel with the generating function: G(y 1 ;:::;y J )= ψ 2C ff y! (7) where is the nest index, and ff are odel paraeters such that 0» ff» 1 8; ; (8) and Vovsha (1997) iposes also that ff > 0 8i: (9) ff i =1 8i: (10) 3

6 Ben-Akiva and Bierlaire (1999) ention the CNL as an instance of a GEV odel, and based on the following generating function: G(y 1 ;:::;y J )= ψ 2C ff y! : (11) A siilar forulation is used by Papola (2000), based on the following generating function: G(y 1 ;:::;y J )= k ψ 2Ck! k 0 ff 0= k ik e V i= k ; (12) with 0» k» 0. Papola iposes also that 2 Theoretical analysis k ff ik =1 8i: (13) Aong these forulations, (11) is the ost general. Indeed, Vovsha's and Sall's forulations are specific cases of (11). We obtain Sall's forulation (4) with = 1 and = 1=ρ. Vovsha's forulation (7) is obtained fro (11) with =1for all. Papola's odel (12) is equivalent to (11), with = 1= 0, = 1= and ff = ff 0=. However, Papola's constraint (13) is not required by our forulation. Note that Sall (5) and Vovsha (10) ipose the sae constraint. The following theore shows that (11) is indeed a GEV generating function. Theore 1 The following conditions are sufficient for (11) to define a GEV generating function: 1. ff 0, 8;, 2. P ff > 0, 8, 3. >0, 4. > 0, 8, 5.», 8. 4

7 Proof. We show that, under these assuptions, (11) verifies the four properties of GEV generating functions. 1. G is obviously non negative, if y 2 R n G is hoogeneous of degree. Indeed, G(fiy) = = = ψ 2C ψ = fi = fi G(y): ff fi y fi 2C ff y ψ fi ff y 2C ψ ff y 2C!!!! 3. The liit properties hold fro assuption 2, that guarantees that there is at least one non zero coefficient ff for each alternative. li yi!1 G(y 1 ;:::;y J ) = li yi!1 = P = 1 P li yi!1 P 2C ff y P 2C ff y (14) 4. The condition for the sign of the derivatives is obtained fro (21) in Lea 2 (see Section 6). We distinguish three cases, considering only y 0. (a) If k =1,we = ff y 1 A 0: (15) (b) If k>1 and =, we k G(y)=@y i1 :::@y ik =0: (16) 5

8 Indeed, k 1 Y n=0 contains a zero factor when n =1. n (17) (c) If k > 1 and <, the sign of (21) is entirely deterined by the sign of (17). For n > 0, we have n < 0 (assuption 5). Therefore, there are k 1 negative and one positive factors in the product. We obtain that k 1 Y n n=0 ( 0 if k is odd» 0 if k is even (18) Therefore, in any case, we have k G(y)=@y i1 :::@y ik 0 if k is odd» 0 if k is even (19) Λ This theore otivates the absence in our forulation of a constraint siilar to (5), (10) and (13). Indeed, such a condition is not required to obtain a valid GEV odel and, consequently, to be consistent with discrete choice theory. Instead, they are used to enable paraeter estiability. Indeed, it is ipossible to estiate all paraeters of the GEV odel, exactly as it is ipossible to estiate all Alternative Specific constants in a MNL or NL odel (see Bierlaire, Lotan and Toint, 1995). 3 Estiation procedure The estiation procedures proposed by Sall (1987) and Vovsha (1997) are based on heuristics. Sall reduces the nuber of free paraeters by iposing arbitrary restrictions on the paraeters: w = 1 M+1, 8, andρ r = ρ, 8r. Vovsha proposes a coplicated heuristic, where each observation is artificially substituted with n observations (Vovsha proposes n =100). We prefer to use a pure axiu likelihood procedure. However, the proble of estiability reains open. A detailed theoretical analysis of odel overspecification, siilar to what Bierlaire et al. (1995) has done for the Alternative Specific 6

9 Constants in the Nested Logit odel, is currently ongoing. Meanwhile, we have developed an optiization process which is able to handle overspecified odels. A new odel estiation package called Biogee (BIerlaire's Optiization routines for GEv Model Estiation) has been developed. It is designed to estiate a wide variety of discrete choice odels. Actually, any odel out of the GEV odel faily can be estiated. Moreover, non linear utility functions can be handled. In particular, a specific scale paraeter can be associated with different groups in the saple. The optiization algorith is based on a quasi-newton BFGS ethod in a trust-region fraework (Conn, Gould and Toint, 2000). The trust-region subproble is solved with a conugate gradient ethod, which does not require the solution of the Newton equations. These equations do not have a solution when the approxiation of the hessian atrix is singular (as it is the case with overspecified odels). Instead, the conugate gradient involves only atrix-vector coputations. The Newton ethod is known to be very slow when the obective is not strictly convex at the solution. Also, it stops when a local axiu of the log-likelihood function has been reached. The search for a global axiu is out of the scope of the ethod. Despite these shortcoings, this optiization algorith is sufficiently robust to epirically analyze the structure of the CNL odel. An efficient estiation procedure will be derived later on, when the structural analysis will be coplete. 4 Preliinary epirical analysis We have perfored a preliinary epirical analysis of the odel estiability. Our ain obective is to assess the relevance of conditions (5), (10) and (13) entioned in the literature. We analyze a trivial odel, with 3 alternatives. The utility functions are ust coposed of the ASCs. The saple contains 3 observations, each one corresponding to a different alternative. It is well know that the axiu likelihood estiator for a MNL odel is obtained with all ASC being equal (say to zero). The associated log-likelihood is 3ln3 ß 3: Clearly, using a CNL odel for such a trivial odel cannot iprove the log-likelihood. But we see that the sae log-likelihood can be obtained with various cobinations of paraeters values. 7

10 asc asc asc ff ff ff ff ff ff Table 1: Optial paraeters for the trivial odel We conecture that if ff = i = K; 8i; K > 0; (20) then the interpretation of the ASC is consistent with rando utility theory. Note that condition (5) by Sall, condition (10) by Vovsha and condition (13) by Papola are equivalent to (20), with K =1. If (20) is not verified, we still have a valid odel with a perfect predictive capability. However, the usual interpretation of the ASC paraeters is not relevant anyore. We illustrate that conecture in Table 1, which contains sets of axiu likelihood estiators of the paraeters. The paraeters contained in each colun provide the exact sae log-likelihood ( 3 ln 3). Condition (20) is verified for coluns 1 to 3. We observe that the optial value of the ASC are the sae as the MNL. In colun 4, where the condition is not verified, the values of the ASC are copletely different. We note also that the ff paraeter ay take values larger than 1, and that the value of K in(20) is not necessarily 1 (K=1 in colun 1, and K = 2 in coluns 2 and 3). We have observed siilar behavior with a non trivial odel on a cobined RP/SP data set for ode choice in Switzerland (Bierlaire, Axhausen and Abbay, 2001). In that case, it sees that the regular" paraeters (for cost or tie) are independent fro the validity of 20, while the 8

11 ASCs are clearly affected. We have shown that (20) is not necessary for the CNL to be part of the GEV odel faily. Therefore, it ay be adequate to ignore it. We suggest that condition (20) is violated for estiation purposes, as it provides ore degrees of freedo to the algoriths and avoid to handle linear constraints. But once an estiated odel is obtained, an equivalent odel verifying (20) ust be derived, in order to provide values of the paraeters that can be analyzed in a usual way. We also note that the constant K in (20) does not necessarily need to be one. Foral proofs of these stateents ust of course be provided. Such proofs are straightforward for the trivial odel used for epirical analysis, but not interesting. The proof for the general case is uch ore difficult. This research is currently ongoing. 5 Conclusion and perspectives In this paper, we have provided and analyzed a general forulation of the crossnested logit odel. It generalizes existing forulations in the literature. We have proved that, under ild conditions, the forulation is consistent with GEV odels. Also, wehave perfored preliinary analysis of a condition on the paraeters iposed on forulations published in the literature. We have proved that this condition is not necessary for the theoretical properties of the odel. But it is iportant for practical properties, related to paraeter estiability, and odel coparison. We have identified an extension of this condition that is iportant in order to correctly interpret the estiated paraeters. The condition is therefore desirable for practical purposes, but not necessary to have a odel with prediction capabilities. The CNL odel is appealing to capture coplex situations where correlations cannot be handled by the Nested Logit odel. Even with few alternative and nests, the use of a CNL instead of a NL odel ay significantly iprove the estiated odel (Bierlaire et al., 2001). The price to pay is the difficulty to estiate the odel paraeters. We have adopted a robust ethod with does not require the odel to be fully estiable in order to find an optial value of the paraeters. However, because the second derivative atrix of the log-likelihood function is singular at the solution, we cannot copute a reliable estiation of the 9

12 variance covariance atrix associated with the estiates. Also, the estiation algorith, based on a quasi-newton approach, can be very slow. An foral analysis of the sources of overspecification, siilar to what Bierlaire et al. (1995) have ade for the Nested Logit Model, is required in order to obtain an efficient estiation process, and a good estiates of the variance-covariance atrix of the estiated paraeters. 6 Annex: Lea The following Lea has been proven by Nicolas Antille, who is gratefully acknowledged. Lea 2 Let i 1,...,i k be k different indices (k > 0) arbitrarily chosen within f1;:::;jg. If G is defined by (11), k G(y)=@y i1 :::@y ik = ψ Yi k k 1 Y! k ffn yn 1 k n A (21) n=i 1 n=0 where A = 2C ff y : (22) Proof. The proofisby induction. We k G(y) = A ff i1 y 1 i1 1 = ff i1 y 1 i 1 A proving the result for k =1. Assuing now that the result is verified for k+1 G(y)=@y i1 :::@y k i1 :::@y ik+1 P P Q = ik k = P k+1 k Q ik n=i1 (ff n y 1 n ) Q k 1 n=0 Q n=i1 (ff n yn 1 k 1 ) Q ik+1 n=i 1 (ff n yn 1 That concludes the proof. Λ n=0 ) Q k n=0 n A n k n A k (k+1) k A : ff ik+1 y 1 i k+1 10

13 7 Annex: Derivatives We provide here the derivatives of the log-likelihood function for GEV odels in general, and for the Cross-Nested Logit odel in particular. Given a saple of observations, the log-likelihood of the saple is L = n2saple ln P (i n C n ); (23) where i n is the alternative actually chosen by individual n, C n is the choice set, and ln P (i n C) = n V in + ln G in (e nv 1 ;:::;e nv J ) (24) ln P e nv G (e nv 1 ;:::;e nv J ) where G i i and n is a scale paraeter associated to individual n. This paraeter allows to estiate odels with heterogeneous saples, without using coplicated nested structures. If fi k is a paraeter appearing in the utility functions V 1 ;:::;V k ln P i P + 1 J G i e k P k G + P n k (25) where = e V G (26) Note that we do not assue here that the V are linear-in-paraeters, so =@fi k is not necessarily a constant. The derivatives with respect to odel paraeters ff k` are k` ln P = 1 G k` 1 k` (27) The derivatives with respect to odel paraeters k are k ln P = 1 G k 1 The derivative with respect to odel paraeter is ln P = 1 G k (28) (29)

14 The derivative with respect to the scale paraeter is ln P = V i + i G 1 (30) = V e (31) V e V G + (32) Finally, we provide the first and second derivatives of (11) with respect to every paraeter. The first derivative with respect to a variable x i is given by! 1 (33) G i = ff i x 1 i ψ ff x The first derivative with respect to the = 1 y ln(y ) (34) where y = ff x (35) 2C The first derivative with respect to the nest paraeter = y ψ and with respect to the ff paraeter is where 2 2 i 2C ff x ln(x = i 1! y 2 ln(y ) (36) x k i (37) y = ff x (38) 2C We now provide the second derivative with respect to x i and x. If i =, we i = y 2 ff i x 2 i (( 1)ff i x i +y ( 1)) (39) 12

15 and if i 6=, we 2 = ( 1)ff i ff y 2 x 1 i x 1 (40) where y = ff x (41) 2C The second derivative withrespect to x i and 2 = y 1 ff i x 1 i (1 + ln(y )) (42) where y = ff x (43) 2C The second derivative withrespect to x i and 2 = y 1 + y 1 ff i x 1 i ff i x 1 i 2 + y y The second derivative withrespect to x i and ff ik 2 G = x k 1 i ik 1 k k ln(y )ff i x 1 i ff i x 1 i ln(x i ): 1+ff ik ( 1)y 1 k x k i k (44) (45) and with respect to x i and ff k (i 6= ) 2 G = ff ik x k 1 i ( k k 2 k k x k (46) where y = ff x (47) 2C References Ben-Akiva, M. and Bierlaire, M. (1999). Discrete choice ethods and their applications to short-ter travel decisions, in R. Hall (ed.), Handbook of Transportation Science, Kluwer, pp

16 Bierlaire, M., Axhausen, K. and Abbay, G. (2001). Acceptance of odel innovation: the case of the swissetro, Proceedings of the 1st Swiss Transport Research Conference. Bierlaire, M., Lotan, T. and Toint, P. L. (1995). On the overspecification of ultinoial and nested logit odels due to alternative specific constants, Technical Report 95/12, Departent of Matheatics, FUNDP, Naur, Belgiu. Conn, A., Gould, N. and Toint, P. (2000). Trust region ethods, MPS SIAM Series on Optiization, SIAM. Forinash, C. V. and Koppelan, F. S. (1993). Application and interpretation of nested logit odels of intercityodechoice, Transportation Research Record Issue 1413: McFadden, D. (1978). Modelling the choice of residential location, in A. K. et al. (ed.), Spatial interaction theory and residential location, North-Holland, Asterda, pp Papola, A. (2000). Soe developent of the cross-nested logit odel, Proceedings of the 9th IATBR Conference. Sall, K. (1987). A discrete choice odel for ordered alternatives, Econoetrica 55(2): Vovsha, P. (1997). Cross-nested logit odel: an application to ode choice in the Tel-Aviv etropolitan area, Transportation Research Board, 76th Annual Meeting, Washington DC. Paper # Vovsha, P. and Bekhor, S. (1998). The link-nested logit odel of route choice: overcoing the route overlapping proble, Transportation Research Record 1645: Willias, H. (1977). On the foration of travel deand odels and econoic easures of user benefit, Environent and Planning 9A:

The Network GEV model

The Network GEV model The Network GEV model Michel Bierlaire January, 2002 Discrete choice models play a major role in transportation demand analysis. Their nice and strong theoretical properties, and their flexibility to capture

More information

Support Vector Machine Classification of Uncertain and Imbalanced data using Robust Optimization

Support Vector Machine Classification of Uncertain and Imbalanced data using Robust Optimization Recent Researches in Coputer Science Support Vector Machine Classification of Uncertain and Ibalanced data using Robust Optiization RAGHAV PAT, THEODORE B. TRAFALIS, KASH BARKER School of Industrial Engineering

More information

Algorithms for parallel processor scheduling with distinct due windows and unit-time jobs

Algorithms for parallel processor scheduling with distinct due windows and unit-time jobs BULLETIN OF THE POLISH ACADEMY OF SCIENCES TECHNICAL SCIENCES Vol. 57, No. 3, 2009 Algoriths for parallel processor scheduling with distinct due windows and unit-tie obs A. JANIAK 1, W.A. JANIAK 2, and

More information

A Simple Regression Problem

A Simple Regression Problem A Siple Regression Proble R. M. Castro March 23, 2 In this brief note a siple regression proble will be introduced, illustrating clearly the bias-variance tradeoff. Let Y i f(x i ) + W i, i,..., n, where

More information

Probability Distributions

Probability Distributions Probability Distributions In Chapter, we ephasized the central role played by probability theory in the solution of pattern recognition probles. We turn now to an exploration of soe particular exaples

More information

Experimental Design For Model Discrimination And Precise Parameter Estimation In WDS Analysis

Experimental Design For Model Discrimination And Precise Parameter Estimation In WDS Analysis City University of New York (CUNY) CUNY Acadeic Works International Conference on Hydroinforatics 8-1-2014 Experiental Design For Model Discriination And Precise Paraeter Estiation In WDS Analysis Giovanna

More information

A Self-Organizing Model for Logical Regression Jerry Farlow 1 University of Maine. (1900 words)

A Self-Organizing Model for Logical Regression Jerry Farlow 1 University of Maine. (1900 words) 1 A Self-Organizing Model for Logical Regression Jerry Farlow 1 University of Maine (1900 words) Contact: Jerry Farlow Dept of Matheatics Univeristy of Maine Orono, ME 04469 Tel (07) 866-3540 Eail: farlow@ath.uaine.edu

More information

E0 370 Statistical Learning Theory Lecture 6 (Aug 30, 2011) Margin Analysis

E0 370 Statistical Learning Theory Lecture 6 (Aug 30, 2011) Margin Analysis E0 370 tatistical Learning Theory Lecture 6 (Aug 30, 20) Margin Analysis Lecturer: hivani Agarwal cribe: Narasihan R Introduction In the last few lectures we have seen how to obtain high confidence bounds

More information

Soft Computing Techniques Help Assign Weights to Different Factors in Vulnerability Analysis

Soft Computing Techniques Help Assign Weights to Different Factors in Vulnerability Analysis Soft Coputing Techniques Help Assign Weights to Different Factors in Vulnerability Analysis Beverly Rivera 1,2, Irbis Gallegos 1, and Vladik Kreinovich 2 1 Regional Cyber and Energy Security Center RCES

More information

e-companion ONLY AVAILABLE IN ELECTRONIC FORM

e-companion ONLY AVAILABLE IN ELECTRONIC FORM OPERATIONS RESEARCH doi 10.1287/opre.1070.0427ec pp. ec1 ec5 e-copanion ONLY AVAILABLE IN ELECTRONIC FORM infors 07 INFORMS Electronic Copanion A Learning Approach for Interactive Marketing to a Custoer

More information

SPECTRUM sensing is a core concept of cognitive radio

SPECTRUM sensing is a core concept of cognitive radio World Acadey of Science, Engineering and Technology International Journal of Electronics and Counication Engineering Vol:6, o:2, 202 Efficient Detection Using Sequential Probability Ratio Test in Mobile

More information

Stochastic Subgradient Methods

Stochastic Subgradient Methods Stochastic Subgradient Methods Lingjie Weng Yutian Chen Bren School of Inforation and Coputer Science University of California, Irvine {wengl, yutianc}@ics.uci.edu Abstract Stochastic subgradient ethods

More information

13.2 Fully Polynomial Randomized Approximation Scheme for Permanent of Random 0-1 Matrices

13.2 Fully Polynomial Randomized Approximation Scheme for Permanent of Random 0-1 Matrices CS71 Randoness & Coputation Spring 018 Instructor: Alistair Sinclair Lecture 13: February 7 Disclaier: These notes have not been subjected to the usual scrutiny accorded to foral publications. They ay

More information

Bipartite subgraphs and the smallest eigenvalue

Bipartite subgraphs and the smallest eigenvalue Bipartite subgraphs and the sallest eigenvalue Noga Alon Benny Sudaov Abstract Two results dealing with the relation between the sallest eigenvalue of a graph and its bipartite subgraphs are obtained.

More information

Extension of CSRSM for the Parametric Study of the Face Stability of Pressurized Tunnels

Extension of CSRSM for the Parametric Study of the Face Stability of Pressurized Tunnels Extension of CSRSM for the Paraetric Study of the Face Stability of Pressurized Tunnels Guilhe Mollon 1, Daniel Dias 2, and Abdul-Haid Soubra 3, M.ASCE 1 LGCIE, INSA Lyon, Université de Lyon, Doaine scientifique

More information

Keywords: Estimator, Bias, Mean-squared error, normality, generalized Pareto distribution

Keywords: Estimator, Bias, Mean-squared error, normality, generalized Pareto distribution Testing approxiate norality of an estiator using the estiated MSE and bias with an application to the shape paraeter of the generalized Pareto distribution J. Martin van Zyl Abstract In this work the norality

More information

Model Fitting. CURM Background Material, Fall 2014 Dr. Doreen De Leon

Model Fitting. CURM Background Material, Fall 2014 Dr. Doreen De Leon Model Fitting CURM Background Material, Fall 014 Dr. Doreen De Leon 1 Introduction Given a set of data points, we often want to fit a selected odel or type to the data (e.g., we suspect an exponential

More information

TEST OF HOMOGENEITY OF PARALLEL SAMPLES FROM LOGNORMAL POPULATIONS WITH UNEQUAL VARIANCES

TEST OF HOMOGENEITY OF PARALLEL SAMPLES FROM LOGNORMAL POPULATIONS WITH UNEQUAL VARIANCES TEST OF HOMOGENEITY OF PARALLEL SAMPLES FROM LOGNORMAL POPULATIONS WITH UNEQUAL VARIANCES S. E. Ahed, R. J. Tokins and A. I. Volodin Departent of Matheatics and Statistics University of Regina Regina,

More information

Testing equality of variances for multiple univariate normal populations

Testing equality of variances for multiple univariate normal populations University of Wollongong Research Online Centre for Statistical & Survey Methodology Working Paper Series Faculty of Engineering and Inforation Sciences 0 esting equality of variances for ultiple univariate

More information

Inspection; structural health monitoring; reliability; Bayesian analysis; updating; decision analysis; value of information

Inspection; structural health monitoring; reliability; Bayesian analysis; updating; decision analysis; value of information Cite as: Straub D. (2014). Value of inforation analysis with structural reliability ethods. Structural Safety, 49: 75-86. Value of Inforation Analysis with Structural Reliability Methods Daniel Straub

More information

Convex Programming for Scheduling Unrelated Parallel Machines

Convex Programming for Scheduling Unrelated Parallel Machines Convex Prograing for Scheduling Unrelated Parallel Machines Yossi Azar Air Epstein Abstract We consider the classical proble of scheduling parallel unrelated achines. Each job is to be processed by exactly

More information

Non-Parametric Non-Line-of-Sight Identification 1

Non-Parametric Non-Line-of-Sight Identification 1 Non-Paraetric Non-Line-of-Sight Identification Sinan Gezici, Hisashi Kobayashi and H. Vincent Poor Departent of Electrical Engineering School of Engineering and Applied Science Princeton University, Princeton,

More information

Revealed Preference with Stochastic Demand Correspondence

Revealed Preference with Stochastic Demand Correspondence Revealed Preference with Stochastic Deand Correspondence Indraneel Dasgupta School of Econoics, University of Nottingha, Nottingha NG7 2RD, UK. E-ail: indraneel.dasgupta@nottingha.ac.uk Prasanta K. Pattanaik

More information

are equal to zero, where, q = p 1. For each gene j, the pairwise null and alternative hypotheses are,

are equal to zero, where, q = p 1. For each gene j, the pairwise null and alternative hypotheses are, Page of 8 Suppleentary Materials: A ultiple testing procedure for ulti-diensional pairwise coparisons with application to gene expression studies Anjana Grandhi, Wenge Guo, Shyaal D. Peddada S Notations

More information

Ch 12: Variations on Backpropagation

Ch 12: Variations on Backpropagation Ch 2: Variations on Backpropagation The basic backpropagation algorith is too slow for ost practical applications. It ay take days or weeks of coputer tie. We deonstrate why the backpropagation algorith

More information

Support recovery in compressed sensing: An estimation theoretic approach

Support recovery in compressed sensing: An estimation theoretic approach Support recovery in copressed sensing: An estiation theoretic approach Ain Karbasi, Ali Horati, Soheil Mohajer, Martin Vetterli School of Coputer and Counication Sciences École Polytechnique Fédérale de

More information

On Poset Merging. 1 Introduction. Peter Chen Guoli Ding Steve Seiden. Keywords: Merging, Partial Order, Lower Bounds. AMS Classification: 68W40

On Poset Merging. 1 Introduction. Peter Chen Guoli Ding Steve Seiden. Keywords: Merging, Partial Order, Lower Bounds. AMS Classification: 68W40 On Poset Merging Peter Chen Guoli Ding Steve Seiden Abstract We consider the follow poset erging proble: Let X and Y be two subsets of a partially ordered set S. Given coplete inforation about the ordering

More information

Bootstrapping Dependent Data

Bootstrapping Dependent Data Bootstrapping Dependent Data One of the key issues confronting bootstrap resapling approxiations is how to deal with dependent data. Consider a sequence fx t g n t= of dependent rando variables. Clearly

More information

A Note on the Applied Use of MDL Approximations

A Note on the Applied Use of MDL Approximations A Note on the Applied Use of MDL Approxiations Daniel J. Navarro Departent of Psychology Ohio State University Abstract An applied proble is discussed in which two nested psychological odels of retention

More information

CS Lecture 13. More Maximum Likelihood

CS Lecture 13. More Maximum Likelihood CS 6347 Lecture 13 More Maxiu Likelihood Recap Last tie: Introduction to axiu likelihood estiation MLE for Bayesian networks Optial CPTs correspond to epirical counts Today: MLE for CRFs 2 Maxiu Likelihood

More information

Revealed Preference and Stochastic Demand Correspondence: A Unified Theory

Revealed Preference and Stochastic Demand Correspondence: A Unified Theory Revealed Preference and Stochastic Deand Correspondence: A Unified Theory Indraneel Dasgupta School of Econoics, University of Nottingha, Nottingha NG7 2RD, UK. E-ail: indraneel.dasgupta@nottingha.ac.uk

More information

LONG-TERM PREDICTIVE VALUE INTERVAL WITH THE FUZZY TIME SERIES

LONG-TERM PREDICTIVE VALUE INTERVAL WITH THE FUZZY TIME SERIES Journal of Marine Science and Technology, Vol 19, No 5, pp 509-513 (2011) 509 LONG-TERM PREDICTIVE VALUE INTERVAL WITH THE FUZZY TIME SERIES Ming-Tao Chou* Key words: fuzzy tie series, fuzzy forecasting,

More information

Sharp Time Data Tradeoffs for Linear Inverse Problems

Sharp Time Data Tradeoffs for Linear Inverse Problems Sharp Tie Data Tradeoffs for Linear Inverse Probles Saet Oyak Benjain Recht Mahdi Soltanolkotabi January 016 Abstract In this paper we characterize sharp tie-data tradeoffs for optiization probles used

More information

Intelligent Systems: Reasoning and Recognition. Perceptrons and Support Vector Machines

Intelligent Systems: Reasoning and Recognition. Perceptrons and Support Vector Machines Intelligent Systes: Reasoning and Recognition Jaes L. Crowley osig 1 Winter Seester 2018 Lesson 6 27 February 2018 Outline Perceptrons and Support Vector achines Notation...2 Linear odels...3 Lines, Planes

More information

Lower Bounds for Quantized Matrix Completion

Lower Bounds for Quantized Matrix Completion Lower Bounds for Quantized Matrix Copletion Mary Wootters and Yaniv Plan Departent of Matheatics University of Michigan Ann Arbor, MI Eail: wootters, yplan}@uich.edu Mark A. Davenport School of Elec. &

More information

Symbolic Analysis as Universal Tool for Deriving Properties of Non-linear Algorithms Case study of EM Algorithm

Symbolic Analysis as Universal Tool for Deriving Properties of Non-linear Algorithms Case study of EM Algorithm Acta Polytechnica Hungarica Vol., No., 04 Sybolic Analysis as Universal Tool for Deriving Properties of Non-linear Algoriths Case study of EM Algorith Vladiir Mladenović, Miroslav Lutovac, Dana Porrat

More information

. The univariate situation. It is well-known for a long tie that denoinators of Pade approxiants can be considered as orthogonal polynoials with respe

. The univariate situation. It is well-known for a long tie that denoinators of Pade approxiants can be considered as orthogonal polynoials with respe PROPERTIES OF MULTIVARIATE HOMOGENEOUS ORTHOGONAL POLYNOMIALS Brahi Benouahane y Annie Cuyt? Keywords Abstract It is well-known that the denoinators of Pade approxiants can be considered as orthogonal

More information

Kernel Methods and Support Vector Machines

Kernel Methods and Support Vector Machines Intelligent Systes: Reasoning and Recognition Jaes L. Crowley ENSIAG 2 / osig 1 Second Seester 2012/2013 Lesson 20 2 ay 2013 Kernel ethods and Support Vector achines Contents Kernel Functions...2 Quadratic

More information

Testing Properties of Collections of Distributions

Testing Properties of Collections of Distributions Testing Properties of Collections of Distributions Reut Levi Dana Ron Ronitt Rubinfeld April 9, 0 Abstract We propose a fraework for studying property testing of collections of distributions, where the

More information

Analyzing Simulation Results

Analyzing Simulation Results Analyzing Siulation Results Dr. John Mellor-Cruey Departent of Coputer Science Rice University johnc@cs.rice.edu COMP 528 Lecture 20 31 March 2005 Topics for Today Model verification Model validation Transient

More information

Normalization and correlation of cross-nested logit models

Normalization and correlation of cross-nested logit models Normalization and correlation of cross-nested logit models E. Abbe, M. Bierlaire, T. Toledo Lab. for Information and Decision Systems, Massachusetts Institute of Technology Inst. of Mathematics, Ecole

More information

Block designs and statistics

Block designs and statistics Bloc designs and statistics Notes for Math 447 May 3, 2011 The ain paraeters of a bloc design are nuber of varieties v, bloc size, nuber of blocs b. A design is built on a set of v eleents. Each eleent

More information

Lost-Sales Problems with Stochastic Lead Times: Convexity Results for Base-Stock Policies

Lost-Sales Problems with Stochastic Lead Times: Convexity Results for Base-Stock Policies OPERATIONS RESEARCH Vol. 52, No. 5, Septeber October 2004, pp. 795 803 issn 0030-364X eissn 1526-5463 04 5205 0795 infors doi 10.1287/opre.1040.0130 2004 INFORMS TECHNICAL NOTE Lost-Sales Probles with

More information

In this chapter, we consider several graph-theoretic and probabilistic models

In this chapter, we consider several graph-theoretic and probabilistic models THREE ONE GRAPH-THEORETIC AND STATISTICAL MODELS 3.1 INTRODUCTION In this chapter, we consider several graph-theoretic and probabilistic odels for a social network, which we do under different assuptions

More information

Department of Electronic and Optical Engineering, Ordnance Engineering College, Shijiazhuang, , China

Department of Electronic and Optical Engineering, Ordnance Engineering College, Shijiazhuang, , China 6th International Conference on Machinery, Materials, Environent, Biotechnology and Coputer (MMEBC 06) Solving Multi-Sensor Multi-Target Assignent Proble Based on Copositive Cobat Efficiency and QPSO Algorith

More information

Normalization and correlation of cross-nested logit models

Normalization and correlation of cross-nested logit models Normalization and correlation of cross-nested logit models E. Abbe M. Bierlaire T. Toledo August 5, 25 Abstract We address two fundamental issues associated with the use of cross-nested logit models. On

More information

Convexity-Based Optimization for Power-Delay Tradeoff using Transistor Sizing

Convexity-Based Optimization for Power-Delay Tradeoff using Transistor Sizing Convexity-Based Optiization for Power-Delay Tradeoff using Transistor Sizing Mahesh Ketkar, and Sachin S. Sapatnekar Departent of Electrical and Coputer Engineering University of Minnesota, Minneapolis,

More information

Interactive Markov Models of Evolutionary Algorithms

Interactive Markov Models of Evolutionary Algorithms Cleveland State University EngagedScholarship@CSU Electrical Engineering & Coputer Science Faculty Publications Electrical Engineering & Coputer Science Departent 2015 Interactive Markov Models of Evolutionary

More information

ESTIMATING AND FORMING CONFIDENCE INTERVALS FOR EXTREMA OF RANDOM POLYNOMIALS. A Thesis. Presented to. The Faculty of the Department of Mathematics

ESTIMATING AND FORMING CONFIDENCE INTERVALS FOR EXTREMA OF RANDOM POLYNOMIALS. A Thesis. Presented to. The Faculty of the Department of Mathematics ESTIMATING AND FORMING CONFIDENCE INTERVALS FOR EXTREMA OF RANDOM POLYNOMIALS A Thesis Presented to The Faculty of the Departent of Matheatics San Jose State University In Partial Fulfillent of the Requireents

More information

Ufuk Demirci* and Feza Kerestecioglu**

Ufuk Demirci* and Feza Kerestecioglu** 1 INDIRECT ADAPTIVE CONTROL OF MISSILES Ufuk Deirci* and Feza Kerestecioglu** *Turkish Navy Guided Missile Test Station, Beykoz, Istanbul, TURKEY **Departent of Electrical and Electronics Engineering,

More information

Recovering Data from Underdetermined Quadratic Measurements (CS 229a Project: Final Writeup)

Recovering Data from Underdetermined Quadratic Measurements (CS 229a Project: Final Writeup) Recovering Data fro Underdeterined Quadratic Measureents (CS 229a Project: Final Writeup) Mahdi Soltanolkotabi Deceber 16, 2011 1 Introduction Data that arises fro engineering applications often contains

More information

Pattern Recognition and Machine Learning. Artificial Neural networks

Pattern Recognition and Machine Learning. Artificial Neural networks Pattern Recognition and Machine Learning Jaes L. Crowley ENSIMAG 3 - MMIS Fall Seester 2016 Lessons 7 14 Dec 2016 Outline Artificial Neural networks Notation...2 1. Introduction...3... 3 The Artificial

More information

Lecture 21. Interior Point Methods Setup and Algorithm

Lecture 21. Interior Point Methods Setup and Algorithm Lecture 21 Interior Point Methods In 1984, Kararkar introduced a new weakly polynoial tie algorith for solving LPs [Kar84a], [Kar84b]. His algorith was theoretically faster than the ellipsoid ethod and

More information

W-BASED VS LATENT VARIABLES SPATIAL AUTOREGRESSIVE MODELS: EVIDENCE FROM MONTE CARLO SIMULATIONS

W-BASED VS LATENT VARIABLES SPATIAL AUTOREGRESSIVE MODELS: EVIDENCE FROM MONTE CARLO SIMULATIONS W-BASED VS LATENT VARIABLES SPATIAL AUTOREGRESSIVE MODELS: EVIDENCE FROM MONTE CARLO SIMULATIONS. Introduction When it coes to applying econoetric odels to analyze georeferenced data, researchers are well

More information

EMPIRICAL COMPLEXITY ANALYSIS OF A MILP-APPROACH FOR OPTIMIZATION OF HYBRID SYSTEMS

EMPIRICAL COMPLEXITY ANALYSIS OF A MILP-APPROACH FOR OPTIMIZATION OF HYBRID SYSTEMS EMPIRICAL COMPLEXITY ANALYSIS OF A MILP-APPROACH FOR OPTIMIZATION OF HYBRID SYSTEMS Jochen Till, Sebastian Engell, Sebastian Panek, and Olaf Stursberg Process Control Lab (CT-AST), University of Dortund,

More information

arxiv: v1 [cs.ds] 3 Feb 2014

arxiv: v1 [cs.ds] 3 Feb 2014 arxiv:40.043v [cs.ds] 3 Feb 04 A Bound on the Expected Optiality of Rando Feasible Solutions to Cobinatorial Optiization Probles Evan A. Sultani The Johns Hopins University APL evan@sultani.co http://www.sultani.co/

More information

Using EM To Estimate A Probablity Density With A Mixture Of Gaussians

Using EM To Estimate A Probablity Density With A Mixture Of Gaussians Using EM To Estiate A Probablity Density With A Mixture Of Gaussians Aaron A. D Souza adsouza@usc.edu Introduction The proble we are trying to address in this note is siple. Given a set of data points

More information

COS 424: Interacting with Data. Written Exercises

COS 424: Interacting with Data. Written Exercises COS 424: Interacting with Data Hoework #4 Spring 2007 Regression Due: Wednesday, April 18 Written Exercises See the course website for iportant inforation about collaboration and late policies, as well

More information

Statistical clustering and Mineral Spectral Unmixing in Aviris Hyperspectral Image of Cuprite, NV

Statistical clustering and Mineral Spectral Unmixing in Aviris Hyperspectral Image of Cuprite, NV CS229 REPORT, DECEMBER 05 1 Statistical clustering and Mineral Spectral Unixing in Aviris Hyperspectral Iage of Cuprite, NV Mario Parente, Argyris Zynis I. INTRODUCTION Hyperspectral Iaging is a technique

More information

Chapter 6 1-D Continuous Groups

Chapter 6 1-D Continuous Groups Chapter 6 1-D Continuous Groups Continuous groups consist of group eleents labelled by one or ore continuous variables, say a 1, a 2,, a r, where each variable has a well- defined range. This chapter explores:

More information

The proofs of Theorem 1-3 are along the lines of Wied and Galeano (2013).

The proofs of Theorem 1-3 are along the lines of Wied and Galeano (2013). A Appendix: Proofs The proofs of Theore 1-3 are along the lines of Wied and Galeano (2013) Proof of Theore 1 Let D[d 1, d 2 ] be the space of càdlàg functions on the interval [d 1, d 2 ] equipped with

More information

Computable Shell Decomposition Bounds

Computable Shell Decomposition Bounds Coputable Shell Decoposition Bounds John Langford TTI-Chicago jcl@cs.cu.edu David McAllester TTI-Chicago dac@autoreason.co Editor: Leslie Pack Kaelbling and David Cohn Abstract Haussler, Kearns, Seung

More information

MODIFICATION OF AN ANALYTICAL MODEL FOR CONTAINER LOADING PROBLEMS

MODIFICATION OF AN ANALYTICAL MODEL FOR CONTAINER LOADING PROBLEMS MODIFICATIO OF A AALYTICAL MODEL FOR COTAIER LOADIG PROBLEMS Reception date: DEC.99 otification to authors: 04 MAR. 2001 Cevriye GECER Departent of Industrial Engineering, University of Gazi 06570 Maltepe,

More information

Inference in the Presence of Likelihood Monotonicity for Polytomous and Logistic Regression

Inference in the Presence of Likelihood Monotonicity for Polytomous and Logistic Regression Advances in Pure Matheatics, 206, 6, 33-34 Published Online April 206 in SciRes. http://www.scirp.org/journal/ap http://dx.doi.org/0.4236/ap.206.65024 Inference in the Presence of Likelihood Monotonicity

More information

Detection and Estimation Theory

Detection and Estimation Theory ESE 54 Detection and Estiation Theory Joseph A. O Sullivan Sauel C. Sachs Professor Electronic Systes and Signals Research Laboratory Electrical and Systes Engineering Washington University 11 Urbauer

More information

Tight Bounds for Maximal Identifiability of Failure Nodes in Boolean Network Tomography

Tight Bounds for Maximal Identifiability of Failure Nodes in Boolean Network Tomography Tight Bounds for axial Identifiability of Failure Nodes in Boolean Network Toography Nicola Galesi Sapienza Università di Roa nicola.galesi@uniroa1.it Fariba Ranjbar Sapienza Università di Roa fariba.ranjbar@uniroa1.it

More information

Fairness via priority scheduling

Fairness via priority scheduling Fairness via priority scheduling Veeraruna Kavitha, N Heachandra and Debayan Das IEOR, IIT Bobay, Mubai, 400076, India vavitha,nh,debayan}@iitbacin Abstract In the context of ulti-agent resource allocation

More information

On the Use of A Priori Information for Sparse Signal Approximations

On the Use of A Priori Information for Sparse Signal Approximations ITS TECHNICAL REPORT NO. 3/4 On the Use of A Priori Inforation for Sparse Signal Approxiations Oscar Divorra Escoda, Lorenzo Granai and Pierre Vandergheynst Signal Processing Institute ITS) Ecole Polytechnique

More information

Quantum algorithms (CO 781, Winter 2008) Prof. Andrew Childs, University of Waterloo LECTURE 15: Unstructured search and spatial search

Quantum algorithms (CO 781, Winter 2008) Prof. Andrew Childs, University of Waterloo LECTURE 15: Unstructured search and spatial search Quantu algoriths (CO 781, Winter 2008) Prof Andrew Childs, University of Waterloo LECTURE 15: Unstructured search and spatial search ow we begin to discuss applications of quantu walks to search algoriths

More information

On the Communication Complexity of Lipschitzian Optimization for the Coordinated Model of Computation

On the Communication Complexity of Lipschitzian Optimization for the Coordinated Model of Computation journal of coplexity 6, 459473 (2000) doi:0.006jco.2000.0544, available online at http:www.idealibrary.co on On the Counication Coplexity of Lipschitzian Optiization for the Coordinated Model of Coputation

More information

A method to determine relative stroke detection efficiencies from multiplicity distributions

A method to determine relative stroke detection efficiencies from multiplicity distributions A ethod to deterine relative stroke detection eiciencies ro ultiplicity distributions Schulz W. and Cuins K. 2. Austrian Lightning Detection and Inoration Syste (ALDIS), Kahlenberger Str.2A, 90 Vienna,

More information

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation Course Notes for EE227C (Spring 2018): Convex Optiization and Approxiation Instructor: Moritz Hardt Eail: hardt+ee227c@berkeley.edu Graduate Instructor: Max Sichowitz Eail: sichow+ee227c@berkeley.edu October

More information

Recursive Algebraic Frisch Scheme: a Particle-Based Approach

Recursive Algebraic Frisch Scheme: a Particle-Based Approach Recursive Algebraic Frisch Schee: a Particle-Based Approach Stefano Massaroli Renato Myagusuku Federico Califano Claudio Melchiorri Atsushi Yaashita Hajie Asaa Departent of Precision Engineering, The University

More information

Moments of the product and ratio of two correlated chi-square variables

Moments of the product and ratio of two correlated chi-square variables Stat Papers 009 50:581 59 DOI 10.1007/s0036-007-0105-0 REGULAR ARTICLE Moents of the product and ratio of two correlated chi-square variables Anwar H. Joarder Received: June 006 / Revised: 8 October 007

More information

Introduction to Discrete Optimization

Introduction to Discrete Optimization Prof. Friedrich Eisenbrand Martin Nieeier Due Date: March 9 9 Discussions: March 9 Introduction to Discrete Optiization Spring 9 s Exercise Consider a school district with I neighborhoods J schools and

More information

Supplementary Information for Design of Bending Multi-Layer Electroactive Polymer Actuators

Supplementary Information for Design of Bending Multi-Layer Electroactive Polymer Actuators Suppleentary Inforation for Design of Bending Multi-Layer Electroactive Polyer Actuators Bavani Balakrisnan, Alek Nacev, and Elisabeth Sela University of Maryland, College Park, Maryland 074 1 Analytical

More information

Feature Extraction Techniques

Feature Extraction Techniques Feature Extraction Techniques Unsupervised Learning II Feature Extraction Unsupervised ethods can also be used to find features which can be useful for categorization. There are unsupervised ethods that

More information

Multi-Dimensional Hegselmann-Krause Dynamics

Multi-Dimensional Hegselmann-Krause Dynamics Multi-Diensional Hegselann-Krause Dynaics A. Nedić Industrial and Enterprise Systes Engineering Dept. University of Illinois Urbana, IL 680 angelia@illinois.edu B. Touri Coordinated Science Laboratory

More information

A note on the multiplication of sparse matrices

A note on the multiplication of sparse matrices Cent. Eur. J. Cop. Sci. 41) 2014 1-11 DOI: 10.2478/s13537-014-0201-x Central European Journal of Coputer Science A note on the ultiplication of sparse atrices Research Article Keivan Borna 12, Sohrab Aboozarkhani

More information

A Better Algorithm For an Ancient Scheduling Problem. David R. Karger Steven J. Phillips Eric Torng. Department of Computer Science

A Better Algorithm For an Ancient Scheduling Problem. David R. Karger Steven J. Phillips Eric Torng. Department of Computer Science A Better Algorith For an Ancient Scheduling Proble David R. Karger Steven J. Phillips Eric Torng Departent of Coputer Science Stanford University Stanford, CA 9435-4 Abstract One of the oldest and siplest

More information

Proc. of the IEEE/OES Seventh Working Conference on Current Measurement Technology UNCERTAINTIES IN SEASONDE CURRENT VELOCITIES

Proc. of the IEEE/OES Seventh Working Conference on Current Measurement Technology UNCERTAINTIES IN SEASONDE CURRENT VELOCITIES Proc. of the IEEE/OES Seventh Working Conference on Current Measureent Technology UNCERTAINTIES IN SEASONDE CURRENT VELOCITIES Belinda Lipa Codar Ocean Sensors 15 La Sandra Way, Portola Valley, CA 98 blipa@pogo.co

More information

RECOVERY OF A DENSITY FROM THE EIGENVALUES OF A NONHOMOGENEOUS MEMBRANE

RECOVERY OF A DENSITY FROM THE EIGENVALUES OF A NONHOMOGENEOUS MEMBRANE Proceedings of ICIPE rd International Conference on Inverse Probles in Engineering: Theory and Practice June -8, 999, Port Ludlow, Washington, USA : RECOVERY OF A DENSITY FROM THE EIGENVALUES OF A NONHOMOGENEOUS

More information

Hybrid System Identification: An SDP Approach

Hybrid System Identification: An SDP Approach 49th IEEE Conference on Decision and Control Deceber 15-17, 2010 Hilton Atlanta Hotel, Atlanta, GA, USA Hybrid Syste Identification: An SDP Approach C Feng, C M Lagoa, N Ozay and M Sznaier Abstract The

More information

Boosting with log-loss

Boosting with log-loss Boosting with log-loss Marco Cusuano-Towner Septeber 2, 202 The proble Suppose we have data exaples {x i, y i ) i =... } for a two-class proble with y i {, }. Let F x) be the predictor function with the

More information

ASSUME a source over an alphabet size m, from which a sequence of n independent samples are drawn. The classical

ASSUME a source over an alphabet size m, from which a sequence of n independent samples are drawn. The classical IEEE TRANSACTIONS ON INFORMATION THEORY Large Alphabet Source Coding using Independent Coponent Analysis Aichai Painsky, Meber, IEEE, Saharon Rosset and Meir Feder, Fellow, IEEE arxiv:67.7v [cs.it] Jul

More information

Pattern Recognition and Machine Learning. Artificial Neural networks

Pattern Recognition and Machine Learning. Artificial Neural networks Pattern Recognition and Machine Learning Jaes L. Crowley ENSIMAG 3 - MMIS Fall Seester 2017 Lessons 7 20 Dec 2017 Outline Artificial Neural networks Notation...2 Introduction...3 Key Equations... 3 Artificial

More information

Topic 5a Introduction to Curve Fitting & Linear Regression

Topic 5a Introduction to Curve Fitting & Linear Regression /7/08 Course Instructor Dr. Rayond C. Rup Oice: A 337 Phone: (95) 747 6958 E ail: rcrup@utep.edu opic 5a Introduction to Curve Fitting & Linear Regression EE 4386/530 Coputational ethods in EE Outline

More information

Generalized eigenfunctions and a Borel Theorem on the Sierpinski Gasket.

Generalized eigenfunctions and a Borel Theorem on the Sierpinski Gasket. Generalized eigenfunctions and a Borel Theore on the Sierpinski Gasket. Kasso A. Okoudjou, Luke G. Rogers, and Robert S. Strichartz May 26, 2006 1 Introduction There is a well developed theory (see [5,

More information

Testing the lag length of vector autoregressive models: A power comparison between portmanteau and Lagrange multiplier tests

Testing the lag length of vector autoregressive models: A power comparison between portmanteau and Lagrange multiplier tests Working Papers 2017-03 Testing the lag length of vector autoregressive odels: A power coparison between portanteau and Lagrange ultiplier tests Raja Ben Hajria National Engineering School, University of

More information

Polygonal Designs: Existence and Construction

Polygonal Designs: Existence and Construction Polygonal Designs: Existence and Construction John Hegean Departent of Matheatics, Stanford University, Stanford, CA 9405 Jeff Langford Departent of Matheatics, Drake University, Des Moines, IA 5011 G

More information

Approximation in Stochastic Scheduling: The Power of LP-Based Priority Policies

Approximation in Stochastic Scheduling: The Power of LP-Based Priority Policies Approxiation in Stochastic Scheduling: The Power of -Based Priority Policies Rolf Möhring, Andreas Schulz, Marc Uetz Setting (A P p stoch, r E( w and (B P p stoch E( w We will assue that the processing

More information

Randomized Recovery for Boolean Compressed Sensing

Randomized Recovery for Boolean Compressed Sensing Randoized Recovery for Boolean Copressed Sensing Mitra Fatei and Martin Vetterli Laboratory of Audiovisual Counication École Polytechnique Fédéral de Lausanne (EPFL) Eail: {itra.fatei, artin.vetterli}@epfl.ch

More information

Use of PSO in Parameter Estimation of Robot Dynamics; Part One: No Need for Parameterization

Use of PSO in Parameter Estimation of Robot Dynamics; Part One: No Need for Parameterization Use of PSO in Paraeter Estiation of Robot Dynaics; Part One: No Need for Paraeterization Hossein Jahandideh, Mehrzad Navar Abstract Offline procedures for estiating paraeters of robot dynaics are practically

More information

Measuring Temperature with a Silicon Diode

Measuring Temperature with a Silicon Diode Measuring Teperature with a Silicon Diode Due to the high sensitivity, nearly linear response, and easy availability, we will use a 1N4148 diode for the teperature transducer in our easureents 10 Analysis

More information

Vulnerability of MRD-Code-Based Universal Secure Error-Correcting Network Codes under Time-Varying Jamming Links

Vulnerability of MRD-Code-Based Universal Secure Error-Correcting Network Codes under Time-Varying Jamming Links Vulnerability of MRD-Code-Based Universal Secure Error-Correcting Network Codes under Tie-Varying Jaing Links Jun Kurihara KDDI R&D Laboratories, Inc 2 5 Ohara, Fujiino, Saitaa, 356 8502 Japan Eail: kurihara@kddilabsjp

More information

paper prepared for the 1996 PTRC Conference, September 2-6, Brunel University, UK ON THE CALIBRATION OF THE GRAVITY MODEL

paper prepared for the 1996 PTRC Conference, September 2-6, Brunel University, UK ON THE CALIBRATION OF THE GRAVITY MODEL paper prepared for the 1996 PTRC Conference, Septeber 2-6, Brunel University, UK ON THE CALIBRATION OF THE GRAVITY MODEL Nanne J. van der Zijpp 1 Transportation and Traffic Engineering Section Delft University

More information

Chaotic Coupled Map Lattices

Chaotic Coupled Map Lattices Chaotic Coupled Map Lattices Author: Dustin Keys Advisors: Dr. Robert Indik, Dr. Kevin Lin 1 Introduction When a syste of chaotic aps is coupled in a way that allows the to share inforation about each

More information

Ştefan ŞTEFĂNESCU * is the minimum global value for the function h (x)

Ştefan ŞTEFĂNESCU * is the minimum global value for the function h (x) 7Applying Nelder Mead s Optiization Algorith APPLYING NELDER MEAD S OPTIMIZATION ALGORITHM FOR MULTIPLE GLOBAL MINIMA Abstract Ştefan ŞTEFĂNESCU * The iterative deterinistic optiization ethod could not

More information

UNCERTAINTIES IN THE APPLICATION OF ATMOSPHERIC AND ALTITUDE CORRECTIONS AS RECOMMENDED IN IEC STANDARDS

UNCERTAINTIES IN THE APPLICATION OF ATMOSPHERIC AND ALTITUDE CORRECTIONS AS RECOMMENDED IN IEC STANDARDS Paper Published on the16th International Syposiu on High Voltage Engineering, Cape Town, South Africa, 2009 UNCERTAINTIES IN THE APPLICATION OF ATMOSPHERIC AND ALTITUDE CORRECTIONS AS RECOMMENDED IN IEC

More information

arxiv: v1 [cs.ds] 17 Mar 2016

arxiv: v1 [cs.ds] 17 Mar 2016 Tight Bounds for Single-Pass Streaing Coplexity of the Set Cover Proble Sepehr Assadi Sanjeev Khanna Yang Li Abstract arxiv:1603.05715v1 [cs.ds] 17 Mar 2016 We resolve the space coplexity of single-pass

More information