Linear programming support vector machines

Size: px
Start display at page:

Download "Linear programming support vector machines"

Transcription

1 Pattern Recognition 35 (00) Linear programming support vector machines Weida Zhou, Li Zhang, Licheng Jiao Key Laboratory for Radar Signal Processing, Xidian University, Xi an, , People s Republic of China Received 3 May 001; accepted 30 October 001 Abstract Based on the analysis of the conclusions in the statistical learning theory, especially the VC dimension of linear functions, linear programming support vector machines (or SVMs) are presented including linear programming linear and nonlinear SVMs. In linear programming SVMs, in order to improve the speed of the training time, the bound of the VC dimension is loosened properly. Simulation results for both articial and real data show that the generalization performance of our method is a good approximation of SVMs and the computation complex is largely reduced by our method.? 00 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved. Keywords: Statistical learning theory; VC dimension; Support vector machines; Generalization performance; Linear programming 1. Introduction Since the 1970s, Vapnik et al. have applied themselves to the study of statistical learning theory [1 3]. Until the early of the 1990, a new kind of learning machines, support vector machine (SVM), was presented based on those theories [,4,5]. The main study of statistical learning theory is the model of learning from examples, which can be described as: there are l random independent identically distributed examples (x 1;y 1); (x ;y );:::;(x l ;y l ); ((x;y) (R n ;R)) drawn according to the uniform probability distribution P(x;y); (P(x;y) = P(x)P(y x)). Given a set of functions f(x;); (where is a parameter set) from which the goal of learning from examples is to select a function f(x; 0) that can express the relationship between x and y in the best possible way. In general, in order to obtain f(x; 0), one has to minimize the expected risk functional R()= L(y; f(x; ))P(x; y)dx dy; (1.1) where L(y; f(x;)) measures the loss between the response y to a given input x and the response f(x;a) provided Corresponding author. Fax: address: zhouwd@rsp.xidian.edu.cn (W. Zhou). by the learning machine. The learning problems such as the problems of pattern recognition, regression estimation and density estimation may be taken as the same learning models withdierent loss functions []. In this paper, we only deal withthe problem of pattern recognition. Consider the following loss function 0ify = f(x;); L(y; f(x;)) = (1.) 1ify f(x;): Due to the probability distribution P(x;y) in Eq. (1.1) is unknown, the expected risk functional is replaced by the empirical risk functional R 1 emp()= L(y i;f(x i;)): (1.3) l In order to know the quality of the empirical risk R emp() to approximate the expected risk, Vapnik presented the following bound theorem []. With probability at least 1 ( ), the inequality ( R() 6 R emp() Remp() ) (1.4) holds true. Where =4(h(ln(l=h) +1) ln )=l and h is the VC dimension of the set of functions f(x;), /0/$.00? 00 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved. PII: S (01)0010-

2 98 W. Zhou et al. / Pattern Recognition 35 (00) From Eq. (1.4), we can see that the minimization of the expected risk R() is equal to the minimization of the two terms on the right-hand side of Eq. (1.4) at the same time. The rst term on the right of Eq. (1.4) R emp() is minimized by learning process. The second term varies with the VC dimension h and the number of examples l. The smaller the VC dimension h and the larger the number of examples l, the smaller the value of the second term. In fact, the number of examples is nite. So for the case of a small example set, the minimization of the expected risk is implemented by minimizing the empirical risk and the VC dimension. Generally speaking, a complex target function set or a large hypothesis space is required for minimizing the empirical risk. But a small hypothesis space is requested for minimizing the VC dimension of the target function set. Therefore, the minimization problem is in a dilemma, the best solution of the problem is to take a compromise between them. Now, restrict the target function to the linear function. Similar to the set of -margin separating hyperplanes de- ned in Ref. [3], we dene the set of m -margin separating hyperplanes. Let us denote the target functions set by f(x; w;b)=w T x + b. If these functions classify an example x as follows: 1; w T x + b ; y = 0 (1.5) 1; w T x + b 6 ; then the set f(x; w;b)=w T x + b is called the set of m -margin separating hyperplanes whose margin is m = w ; (1.6) where denotes l -norm, namely Euclidean distance. There is a conclusion about the VC dimension of the set of m -margin separating hyperplanes. Let vectors x X belong to a sphere of radius R. Then the set of m -margin separating hyperplanes has the VC dimension h bounded by the inequality: R ;n +1; (1.7) m where n is the dimension of input space. From Eq. (1.7), we can see that if the VC dimension of the target functions set h is n, then h varies inversely withthe margin m. In this way, f(x; w 0;b 0) can be approached by minimizing the empirical risk functional and maximizing the separating margin m, which is the structural risk in SVMs introduced by Vapnik: R 1 structure()=cr emp()+ ; (1.8) m where the constant C 0 is a parameter chosen by the users and is a parameter of the target function set. For l random independent identically distributed examples (x 1;y 1); (x ;y );:::;(x l ;y l ); ((x;y) (R n ;R)), the linear SVMs for pattern recognition have the following optimization problem [4]: min 1 w + C i; (1.9) s:t: y i((w x i)+b) 1 i; i 0; i =1;:::;l; where ( ; ) denotes the inner product. Minimizing the rst term in Eq. (1.9) 1 w plays the role of controlling the capacity of the learning machine and avoiding the overtting of the machine. While minimizing the second term is to minimize the empirical risk. The Wolfe dual programming of Eq. (1.9) is [4] max s:t: W ()= i 1 i jy iy j(x i x j); (1.10) i;j=1 iy i =0; (1.11) i [0;C]; i =1;:::;l: (1.1) The decision function has the following form: f(x)= and iy i(x i x)+b (1.13) y = sgn(f(x)): (1.14) The kernel functions are introduced in linear SVMs, which leads to nonlinear SVMs [4]: max s:t: W ()= i 1 i jy iy jk(x i x j) (1.16) i;j=1 iy i = 0 (1.17) i [0;C]; i =1;:::;l: (1.18) And then the decision function can be written as f(x)= iy ik(x i x)+b: (1.19) In a nutshell, the theory foundation of SVMs (statistical learning theory) is rather perfect. But training a SVM requires the solution of a quadratic programming (QP) optimization that is not easy to implement, in particular for large-scale problems. Based on the statistical learning theory, the linear programming SVMs that are extremely simple without explicitly solving QP problems are proposed.

3 W. Zhou et al. / Pattern Recognition 35 (00) Linear programming support vector machines (LPSVMs).1. Linear LPSVMs Lemma.1. [6]: the equivalence of vector norm in n-dimension linear space V n. Let x and x be any vector norms in a nite dimension linear space V n. The norms include l p-norm; weight-norm and others. Then there exist two positive constants 0 c 1 + and 0 c + suchthat c 1 x 6 x 6 c x x V n holds true. Theorem.. Given the examples set ((x 1;y 1);:::;(x i;y i); :::;(x l ;y l )); (x;y) (R n ;R) and the vectors x belong to a sphere of radius R. If the set of m -margin separating hyperplanes f(x; w;b)=w T x + b classify vectors x as follows: 1; w T x + b ; y = 0: 1; w T x + b 6 ; Then there exist a constant 0 c + without depending on examples (x;y) and the parameters w and b suchthat the bound inequality c R w ;n + 1 (.1) holds true. Where w is any vector norm. The proof of Theorem. is shown in Appendix A. The theorem can lead to the following programming directly. Given the examples set ((x 1;y 1);:::;(x i;y i);:::;(x l ;y l )); (x;y) (R n ;R) and the set of linear target functions f(x; w;b)=w T x + b. We introduce the structural risk according to Theorem.. At rst, we will start with the linearly separable case, or without error. max L = r (.) c w s:t: y i(w T x i + b) r i =1;:::;l; r 0; where w is any vector norm and c is a constant. If w is -norm, then c =1. Now consider l -norm x. Certainly there exists a positive constant c without the set of examples and the parameters of the target functions set such that c R w ;n +1 holds on. And so Eq. (.) can be rewritten as r L = : (.3) c w The denominator in Eq. (.) is translated into a constraint. Since the constant c has no eect on solving of linear programming, it can be ignored. In doing so, we have the following programming: LP1: max L = r (.4) s:t: y i(w T x i + b) r i =1;:::;l; (.5) r w j 6 1 j =1;:::;n: The decision function takes the form f(x; w;b)=w T x + b (.6) and the output of pattern recognition is y = sgn(f(x; w;b)); (.7) where sgn( ) is the sign function. For the nonseparable case, we have the following linear programming by introducing positive slack variable i: LP: min L = r + C i (.8) s:t: y i(w T x i + b) r i i =1;:::;l; (.9) r w j 6 1 j =1;:::;n; i 0 i =1;:::;l; where C 0 is a constant. The decision function of the nonseparable case is the same as the separable case... Nonlinear LPSVMs Similar to the nonlinear SVMs [1 4], here we get the nonlinear LPSVMs by adopting the kernel functions K(x i; x j) [7]. In Eqs. (.5) and (.9), the relationship between weights w and vectors x is expressed by the form of dot products, (w x). Now suppose we rst mapped the examples from n-dimensional space into some high-dimensional Euclidean space H: : R n H: Then of course the training algorithm would only depend on the examples through dot products in H, i.e. on functions of the form ((x i) (x j)). If there were a kernel function K suchthat K(x i; x j)=(x i) (x j), we would only need to use K in the training algorithm and would need to know what the mapping function is. Theorem.3. Given the set of examples (((x 1);y 1);:::; ((x i);y i);:::;((x l );y l )); ((x);y) (H; R). The vectors x belong to a sphere of radius R. Let the set of target

4 930 W. Zhou et al. / Pattern Recognition 35 (00) functions be f(x)= iy i(x i) (x)+b = iy ik(x i; x)+b: (.10) And the decision on examples x takes the form 1; f(x) ; y = 0 1; f(x) 6 ; then there exist a constant 0 c + without depending the examples (x;y) and the parameter of the target functions set and b suchthat c R ;n + 1 (.11) holds on. Where =( 1 l ) T and is any vector norm. The proof of Theorem.3 is shown in Appendix B. Similar to linear LPSVMs, consider. We can have the following programming: LP3: max L = r (.1) ( ) s:t: y if(x i)=y i jy jk(x j; x i)+b r j=1 i =1;:::;l; (.13) r k 6 1 k =1;:::;l: The decision function is f(x)= iy ik(x i; x)+b (.14) and the output of pattern recognition is y = sgn(f(x)): (.15) For the nonseparable case, we introduced the slack variables i to get the following linear programming: LP4: min L = r + C i (.16) s:t: y if(x i)=y i ( j=1 jy jk(x j; x i)+b ) r i i =1;:::;l; (.17) r k 6 1 k =1;:::;l; i 0 i =1;:::;l; where C 0 is a constant. The decision function is identical to Eqs. (.14) and (.15). In Section.1, the linear LPSVMs are obtained under the hypothesis of linear target functions set f(x; w;b)=w T x + b. For high-dimensional examples (namely the number of examples is less than or equal to the dimension of example), we can use the target functions f(x)= l iyi(xi x)+b, and then get the following linear LPSVMs: LP5: min L = r + C i (.18) s:t: y if(x i)=y i ( j=1 jy j(x j x i)+b ) r i i =1;:::;l; (.19) r k 6 1 k =1;:::;l; i 0 i =1;:::;l: 3. Simulation experiments Four experiments have been done. For linear LPSVMs, a set of articial data was used in experiments. And for nonlinear LPSVMs, the dual-spiral data, the handwritten digital data and the DS-CDMA multiuser detection data were adopted. For the sake of comparison, we used the programs in Matlab Optimization toolboxes (Version 5.3) for bothquadratic programming support vector machines (or QPSVMs) and LPSVMs The experiment on articial data Let us consider two-class articial linear separable data. 100 training examples and 00 test ones are generated randomly according to the uniform distribution in rectangles ((0; 0); (1= ; 1= ); (0; ); ( 1= ; 1= )) and ((1= ; 1= ); ( ; ); (1= ; 3= ); (0; )). The results are shown in Table 1 and Fig. 1. The results lead us to conclude that the training speed of LPSVMs is faster than that of QPSVMs without loss in the classication performance. 3.. The experiment on the dual-spiral data The dual-spiral problem is referred to as touchstone to test the ability of learning algorithm, which requires us to classify the points on one spiral from ones on another spiral in two-dimensional Cartesian coordinates [8]. In this experiment, we used Gaussian kernel K(x i; x j)= exp( x i x j =p ) with p = 8. The results are shown in Table and Fig.. From the results in Table it can be concluded that the training speed of LPSVMs has been improved remarkably. The VC dimension of LPSVMs is larger than that of QPSVMs, so in theory the generalization performance of LPSVMs is worse than that of QPSVMs,

5 W. Zhou et al. / Pattern Recognition 35 (00) Table 1 The results of articial data Learning No. of training No. of testing Training time Recognition R =m model examples examples (s) error (%) QPSVMs LPSVMs Fig. 1. The results of articial data obtained by LPSVMs. The real line is f(x) = 0, the dotted lines are f(x)=±r. which is not be conrmed by the recognition errors in Table because of the examples without noise. From Fig. we can see that the generalization performance of LPSVMs is worse than that of QPSVMs. At the cost of generalization performance the training speed of LPSVMs is improved. Fortunately the loss in generalization performance is tolerable because the recognition errors is invisible The experiment on the handwritten digits data The experiment data is the MNIST database of 60,000 training and 10,000 testing handwritten digits from AT& T ResearchLabs 1, which has been taken from the experiment data in Refs. [9 11]. Since the database is large, we only had two-class examples belonging to the classes 6 and 9, respectively, and normalized these examples. The kernel function is Gaussian kernel with p = 30. From the results shown in Table 3, we can conclude that the training time of QPSVMs increases exponentially as the increase of the number of examples, while that of LPSVMs increases linearly. Therefore, LPSVMs at least make an order of magnitude improvement in training speed Multiuser detection in DS-CDMA system Code division multiple access (or CDMA) is characterized by its soft capacity, dynamic sharing of channel 1 URL: vann/ocr/moist. resources, high anti-jamming capability and others [1]. The introduction of Multiuser detection technique is to solve the multiple access interference problems (mainly caused by the nonorthogonality of the signature waveform, the asynchronization and the multipath propagation) in wireless communication, which improves the quality and the quantity of communication. Simultaneously, the high complex requirement of signal processing in CDMA systems leads to a wide attention in the eld of signal processing. Multiuser detection is a new direction in CDMA systems [13,14]. Many methods of signal processing have been proposed to solve multiuser detection. But from the view of the - nal goal of multiuser detection, it can be taken as a pattern recognition problem. The goal of multiuser detection is to obtain the minimal bit-error-rate, so the optimal measure of detectors performance is the very principle of the minimum probability of error, not others such as minimum mean-square error, maximum signal-to-noise rate and so on. There exists close relationship or even equivalence between those signal processing principles and the principle of the minimum probability of error, but dierence in many cases. At present, the methods for pattern recognition have received considerable attention in multiuser detection. Here, we take the multiuser detection problem as a fourth experiment. The baseband signal of multiuser detection in a synchronous DS-CDMA system can be expressed as x(t)= K A ib i(j)s i(t jt )+n(t); t [jt; jt + T ]; where T is the symbol interval, b i(j) 1; 1}} is the bit transmitted by the ithuser, A i is the received amplitude 1 of the ithuser, A i is referred to as the energy of the ith user, s i(t); t [0;T] is the deterministic signature waveform assigned to the ithuser and n(t) is white Gaussian noise withunit power spectral density. It is assumed that s i(t) is supported only on the interval [0;T] and has unit energy and that b i(j)} is a collection of independent equiprobable ±1 random variables. We sample the input signal at a rate of 1=T s.iflett s = T=N, where N is the spreading gain, then the sampling rate is equal to the chip rate. On the interval t [jt; jt + T ], we have a vector form x(j)= K A ib i(j)s i + n(j); x(j); s i; n(j) R N 1 :

6 93 W. Zhou et al. / Pattern Recognition 35 (00) Table The results of the dual-spiral data Learning No. of training No. of testing Training time Recognition R =m model examples examples (s) error (%) QPSVMs LPSVMs Fig.. The result of the dual-spiral data obtained by LPSVMs is shown in (a) and QPSVMs in (b). The real lines are f(x) = 0 and the dotted lines are f(x)= ±r. Table 3 The results of the handwritten digital data No. of No. of LPSVMs QPSVMs training testing examples examples Training Recognition R =m Training Recognition R =m time (s) error (%) time (s) error (%) The baseband signal of multiuser detection in an asynchronous DS-CDMA system can be expressed as x(t)= K A ib i(j)s i(t jt i)+n(t); t [jt; jt +T ]; where i is a relative oset, given i 6 6 K. Synchronizing the signal with the expected kth user and sampling at the chip rate, we get K x(j)=a k b k (j)s k + A ib i(j)s i + n(j); i k x(j); s k ; n(j) R N 1 ; where b i and s i denote the equivalent bit stream and the signature waveform of the ithuser, respectively. In general, an

7 W. Zhou et al. / Pattern Recognition 35 (00) asynchronous user can be transformed into two synchronous users. Consider a DS-CDMA system. Signature waveform is Golden code with 31 bits. Let the power of the expected user be denoted by P EU. A relative power P rui is dened by the power ratio of the ithuser to the expected user, namely P rui = P IUi =P EU where P IUi = 1 A i is the power of the ith user and P EU = 1 A k the expected user. Similarly Gaussian noise power is expressed as P N = NP rnp EU, where N is spreading gain and P rn is the relative power of noise. We performed synchronous and asynchronous multiuser detection by using SVM techniques. The results are shown in Tables 4 and 5. The baseband signal of multiuser detection was sampled to generate the examples. Approximately the examples have normal distribution with dierent mean vectors and a common covariance matrix. According to Raudys [15], we randomly took the training examples N = 100 3p, where p denotes the dimension of examples and the test ones 10,000 in each run. Table 4 shows the results of synchronous multiuser detection. The number of users is 10, the power of expected user is 1, the relative power of other users is P ru= =P ru5=10(db) and P ru6= =P ru10=0(db) and the relative power of Gaussian noise is shown in Table 4. The results of asynchronous multiuser detection are given in Table 5. The number of asynchronous users is 6 and each of asynchronous users (including the expected one) has three propagation paths. Let the power of the expected user be 1. The relative power of other users is P ru = P ru3 = 10(db) and P ru4 = P ru5 = P ru6 = 0(db). The power of eachpathis identical for the same user. The relative power of Gaussian noise is indicated in Table 5. In this experiment, the recognition errors of training examples are identical for LPSVMs and QPSVMs. The parameters of RBF are obtained by using leave-one-out methods of QPSVMs [16], so they are optimal for QPSVMs. Experimental results support LPSVMs presented and witness the validity of LPSVMs. The bound of VC dimension in LPSVMs is loosened to some extent, which leads to the loss in generalization performance that is tolerable. Thus, LPSVMs have linear structure risk by relaxing the VC bound and improve the speed of training. 4. Conclusion and discussion Linear programming SVMs are proposed based on the statistical learning theory, in particular the theory of capacity control, VC dimension and structural risk. The bound of VC dimension in LPSVMs is loosened to some extent, which leads to the loss in generalization performance that is tolerable. Training LPSVMs is simpler than QPSVMs, especially for large-scale problems. LPSVMs at least make an order magnitude improvement in the training speed, so it is worthwhile losing some generalization performance. We note that the VC dimension of LPSVMs is larger than that of QPSVMs in Section 3.4, but the generalization error Table 4 Synchronous multiuser detection Gauss Linear LPSVMs Linear QPSVMs RBF LPSVMs RBF QPSVMs noise C =10 C =10 p1 = 406; C =10; 000 p1 = 406; C =10; 000 Prn Training Recognition R =m Training Recognition R =m Training Recognition R =m Training Recognition R =m (db) time (s) error (%) time (s) error (%) time (s) error (%) time (s) error (%)

8 934 W. Zhou et al. / Pattern Recognition 35 (00) Table 5 Asynchronous multiuser detection Gauss Linear LPSVMs Linear QPSVMs RBF LPSVMs RBF QPSVMs noise C = 100 C = 100 p1 = 490; C =10; 000 p1 = 490; C =10; 000 Prn Training Recognition R =m Training Recognition R =m Training Recognition R =m Training Recognition R =m time (s) error (%) time (s) error (%) time (s) error (%) time (s) error (%) of LPSVMs is smaller than that of QPSVMs in linear case, which is not strange. As we know, SVMs dene a structure risk by the empirical risk and the VC dimension and give a bound to the expected risk. Although SVM algorithm can guarantee the structure risk converge to the expected risk withincreasing number of examples. The structure risk is not the expected one exactly. Especially the bound of the expected risk is not very accuracy for the nite number of examples. It is possible that other learning machines get better generalization performance than SVMs, for example, Euclidean distance classier is the optimal one for Gaussian model [15]. Roughly we agree to the viewpoint in Ref. [17]: SVMs are the general and very ecient statistical learning model, but not the optimal. Priori information is helpful for us to select proper learning machine. In fact it is the problem concerned by statistical theory researchers that how to introduce the priori information into SVMs to obtain good generalization performance. Presently in QPSVMs many algorithms for large-scale problems have been developed, such as Chunking algorithm [18], Osuna algorithm [19] and SMO algorithm [0]. Further studies will aim to nd the algorithm for large-scale problems in LPSVMs. In addition the property (say geometry) of LPSVMs deserves further research. Acknowledgements Thanks to all those who contributed to the data sets used in this paper. This work was supported in part by grants of the Nature Science fund (No ) and the National 863 project number Appendix A. Proof of Theorem. Proof. Given the examples set ((x 1;y 1);:::;(x i;y i);:::; (x l ;y l )); (x;y) (R n ;R). The examples belong to a sphere of radius R. The set of m -margin separating hyperplanes f(x; w;b)=w T x + b classies examples x as follows: 1; w T x + b ; y = 0; 1; w T x + b 6 ; according to Eqs. (1.6) and (1.7); then the upper bounds on the VC dimension of the set of m -margin separating hyperplanes R w ;n +1: (A.1) In terms of Lemma.1; there exist 0 c + withno dependence withthe vector w suchthat w 6 c w (A.)

9 W. Zhou et al. / Pattern Recognition 35 (00) holds true. Where w is any vector norm. Hence; we have the upper bounds on the VC dimension R w ;n +1 c R w 6 min ;n +1: Thus Theorem. is proved. Appendix B. Proof of Theorem.3 Proof. Given the examples set (((x 1);y 1);:::;((x i); y i);:::;((x l );y l )); ((x);y) (H; R). The examples belong to a sphere of radius R. The set of the target functions f(x)= = iy i(x i) (x)+b iy ik(x i; x)+b classies examples x as follows: 1; f(x) ; y = 1; f(x + b) 6 ; 0: (B.1) The set of the target functions is identical to a hyperplane in the space H: In terms of Eqs. (1.6) and (1.7); the bounds is R w ;n +1 R T K = min ;n +1: (B.6) In the following we introduce the concept of weight-norm. Denition.3 (Weight-norm [6]). Let A be any nthpositive denite matrix and x R n be row vectors. Then a function x A =(x T Ax) 1= is a kind of vector norm called weight-norm; or ellipse-norm. By the denition of weight-norm, Eq. (5:7) is a weight-norm. We have w = T K = a K: (B.7) According to Lemma.1, there exits a constant 0 c + without depending on and K suchthat K 6 c (B.8) holds true. Where is any vector norm. Combining Eqs. (5:8), (5:9) and (5:10), we have the bounds on the VC dimension: R T K ;n +1 ([ R K = min ] ) ;n +1 c R 6 min ;n +1: Thus, Theorem is proved. f(x)=w (x)+b; (B.) References where w = iy i(x i): In H; according to Eq. (5:5) we have w = w w = = i jy iy j(x i) (x j) j=1 i jy iy jk(x i; x j): j=1 (B.3) (B.4) Let K denote a l l matrix K(i; j) =y iy jk(x i; x j) and =( 1;:::; l ) T. Then Eq. (5:6) can be rewritten as w = a T Ka: (B.5) [1] V. Vapnik, Statistical Learning Theory, Wiley-Interscience Publication, New York, [] V. Vapnik, The Nature of Statistical Learning Theory, Springer, NY, [3] V. Vapnik, An overview of statistical learning theory, IEEE Transac. Neural Networks 10 (5) (1999) [4] C.J.C. Burges, A tutorial on support vector machines for pattern recognition, Data Min. Knowledge Discovery (1998) [5] A.J. Smola, B. Scholkopf, A tutorial on support vector regression, NeuroCOLT Technical Report NC-TR , Royal Holloway College, University of London, UK, [6] W.P. Cheng, K.Y. Zhang, Z. Xu et al. (Eds.), Matrix Theory. Xi an: North-West Industry University Press, Xi an, China, [7] G. Wahba, An introduction to model building with reproducing kernel hilbert spaces, Technical Report 100, University of Wisconsin-Madison, Statistics Dept [8] K.J. Lang, M.J. Witbrock, Learning to tell two spirals apart, Proceedings of the 1989 Connectionist Models Summer School, 1989, pp

10 936 W. Zhou et al. / Pattern Recognition 35 (00) [9] B. Scholkopf, C. Burges, V. Vapnik, Extracting support data for a given task, in: M. Fayyad, R. Uthurusamy (Eds.), Proceedings, First International Conference on Knowledge Discovery & Data Mining, Menlo Park, AAAI Press, [10] B. Scholkopf, C. Burges, V. Vapnik, Incorporating invariances in support vector learning machines, in: C. von der Malsburg, W. von Seelen, J.C. Vorbrggen, B. Sendho (Eds.), Articial Neural Networks ICANN 96, Berlin, Lecture Notes in Computer Science, Vol. 111, Springer, Berlin, 1996, pp [11] C.J.C. Burges, B. Scholkopf, Improving the accuracy and speed of support vector learning machines, in: M. Mozer, M. Jordan, T. Petsche (Eds.), Advances in Neural Information Processing Systems 9, MIT Press, Cambridge, MA, 1997, pp [1] S. Verdu, Multiuser Detection, Cambridge University Press, Cambridge, [13] U. Madhow, Blind adaptive interference suppression for timing acquisition and demodulation in direct-sequence CDMA systems, IEEE Trans. Commun. 46 (1998) [14] X. Wang, H.V. Poor, Blind multiuser detection: a subspace approach, IEEE Trans. Inform. Theory 44 (1998) [15] S. Raudys, Evolution and generalization of a size neurone, II. Complexity of statistical classiers and sample size consideration, Neural Networks 11 (1998) [16] T. Joachims. Estimating the generalization performance of a SVM eciently, LS VIII-Report 5, Universitat Dortmund, Germany. Available in URL: ( [17] S. Raudys, How good are support vector machines, Neural Networks 13 (000) [18] V. Vapnik, Estimation of Dependencies Based on Empirical Data, Nauka, Moscow, 1979 (in Russian) (Englishtranslation: Springer, New York, 198). [19] E. Osuna, R. Freund, G. Girosi, Improved training algorithm for support vector machines, Proceedings of the IEEE NNSP 97, Amelia Island, [0] J. Platt, Fast training of support vector machines using sequential minimal optimization, in: B. Scholkopf, C.J.C. Burges, A.J. Smola (Eds.), Advances in Kernel Methods Support Vector Learning, MIT Press, Cambridge, MA, 1999, pp About the Author WEIDA ZHOU received his B.S. in Engineering from Xidian University, Xi an, China, in Since 1998 he has been working towards the M.S. degree and the Ph.D. degree at Xidian University. His research interests include machine learning, learning theory and data mining. About the Author LI ZHANG received her B.S. degree in Electronic Engineering from Xidian University, Xi an, China. Since 1997 she has been working towards the M.S. degree and the Ph.D. degree at Xidian University. Her research interests have been in the areas of pattern recognition, machine learning and data mining. About the Author LICHENG JIAO received his B.S. from Shanghai Jiaotong University, Shanghai, China, in 198, the M.S. degree and Ph.D. degree from Xi an Jiaotong University, Xi an, China, in 1984 and 1990, respectively. He is currently Professor and Subdecanal of Graduate School. His research interests include neural network, data mining, nonlinear intelligence signal processing and communication.

Support Vector Regression with Automatic Accuracy Control B. Scholkopf y, P. Bartlett, A. Smola y,r.williamson FEIT/RSISE, Australian National University, Canberra, Australia y GMD FIRST, Rudower Chaussee

More information

An introduction to Support Vector Machines

An introduction to Support Vector Machines 1 An introduction to Support Vector Machines Giorgio Valentini DSI - Dipartimento di Scienze dell Informazione Università degli Studi di Milano e-mail: valenti@dsi.unimi.it 2 Outline Linear classifiers

More information

A GENERAL FORMULATION FOR SUPPORT VECTOR MACHINES. Wei Chu, S. Sathiya Keerthi, Chong Jin Ong

A GENERAL FORMULATION FOR SUPPORT VECTOR MACHINES. Wei Chu, S. Sathiya Keerthi, Chong Jin Ong A GENERAL FORMULATION FOR SUPPORT VECTOR MACHINES Wei Chu, S. Sathiya Keerthi, Chong Jin Ong Control Division, Department of Mechanical Engineering, National University of Singapore 0 Kent Ridge Crescent,

More information

Support Vector Machine Regression for Volatile Stock Market Prediction

Support Vector Machine Regression for Volatile Stock Market Prediction Support Vector Machine Regression for Volatile Stock Market Prediction Haiqin Yang, Laiwan Chan, and Irwin King Department of Computer Science and Engineering The Chinese University of Hong Kong Shatin,

More information

Formulation with slack variables

Formulation with slack variables Formulation with slack variables Optimal margin classifier with slack variables and kernel functions described by Support Vector Machine (SVM). min (w,ξ) ½ w 2 + γσξ(i) subject to ξ(i) 0 i, d(i) (w T x(i)

More information

Support Vector Machines vs Multi-Layer. Perceptron in Particle Identication. DIFI, Universita di Genova (I) INFN Sezione di Genova (I) Cambridge (US)

Support Vector Machines vs Multi-Layer. Perceptron in Particle Identication. DIFI, Universita di Genova (I) INFN Sezione di Genova (I) Cambridge (US) Support Vector Machines vs Multi-Layer Perceptron in Particle Identication N.Barabino 1, M.Pallavicini 2, A.Petrolini 1;2, M.Pontil 3;1, A.Verri 4;3 1 DIFI, Universita di Genova (I) 2 INFN Sezione di Genova

More information

Scale-Invariance of Support Vector Machines based on the Triangular Kernel. Abstract

Scale-Invariance of Support Vector Machines based on the Triangular Kernel. Abstract Scale-Invariance of Support Vector Machines based on the Triangular Kernel François Fleuret Hichem Sahbi IMEDIA Research Group INRIA Domaine de Voluceau 78150 Le Chesnay, France Abstract This paper focuses

More information

Support Vector Machines

Support Vector Machines Support Vector Machines Tobias Pohlen Selected Topics in Human Language Technology and Pattern Recognition February 10, 2014 Human Language Technology and Pattern Recognition Lehrstuhl für Informatik 6

More information

Support Vector Machines

Support Vector Machines Support Vector Machines Stephan Dreiseitl University of Applied Sciences Upper Austria at Hagenberg Harvard-MIT Division of Health Sciences and Technology HST.951J: Medical Decision Support Overview Motivation

More information

Support Vector Machines II. CAP 5610: Machine Learning Instructor: Guo-Jun QI

Support Vector Machines II. CAP 5610: Machine Learning Instructor: Guo-Jun QI Support Vector Machines II CAP 5610: Machine Learning Instructor: Guo-Jun QI 1 Outline Linear SVM hard margin Linear SVM soft margin Non-linear SVM Application Linear Support Vector Machine An optimization

More information

Lecture 18: Kernels Risk and Loss Support Vector Regression. Aykut Erdem December 2016 Hacettepe University

Lecture 18: Kernels Risk and Loss Support Vector Regression. Aykut Erdem December 2016 Hacettepe University Lecture 18: Kernels Risk and Loss Support Vector Regression Aykut Erdem December 2016 Hacettepe University Administrative We will have a make-up lecture on next Saturday December 24, 2016 Presentations

More information

A note on the generalization performance of kernel classifiers with margin. Theodoros Evgeniou and Massimiliano Pontil

A note on the generalization performance of kernel classifiers with margin. Theodoros Evgeniou and Massimiliano Pontil MASSACHUSETTS INSTITUTE OF TECHNOLOGY ARTIFICIAL INTELLIGENCE LABORATORY and CENTER FOR BIOLOGICAL AND COMPUTATIONAL LEARNING DEPARTMENT OF BRAIN AND COGNITIVE SCIENCES A.I. Memo No. 68 November 999 C.B.C.L

More information

ARTIFICIAL INTELLIGENCE LABORATORY. and. Comparing Support Vector Machines. with Gaussian Kernels to. F. Girosi, P. Niyogi, T. Poggio, V.

ARTIFICIAL INTELLIGENCE LABORATORY. and. Comparing Support Vector Machines. with Gaussian Kernels to. F. Girosi, P. Niyogi, T. Poggio, V. MASSACHUSETTS INSTITUTE OF TECHNOLOGY ARTIFICIAL INTELLIGENCE LABORATORY and CENTER FOR BIOLOGICAL AND COMPUTATIONAL LEARNING DEPARTMENT OF BRAIN AND COGNITIVE SCIENCES A.I. Memo No. 1599 December, 1996

More information

These outputs can be written in a more convenient form: with y(i) = Hc m (i) n(i) y(i) = (y(i); ; y K (i)) T ; c m (i) = (c m (i); ; c m K(i)) T and n

These outputs can be written in a more convenient form: with y(i) = Hc m (i) n(i) y(i) = (y(i); ; y K (i)) T ; c m (i) = (c m (i); ; c m K(i)) T and n Binary Codes for synchronous DS-CDMA Stefan Bruck, Ulrich Sorger Institute for Network- and Signal Theory Darmstadt University of Technology Merckstr. 25, 6428 Darmstadt, Germany Tel.: 49 65 629, Fax:

More information

Perceptron Revisited: Linear Separators. Support Vector Machines

Perceptron Revisited: Linear Separators. Support Vector Machines Support Vector Machines Perceptron Revisited: Linear Separators Binary classification can be viewed as the task of separating classes in feature space: w T x + b > 0 w T x + b = 0 w T x + b < 0 Department

More information

Cheng Soon Ong & Christian Walder. Canberra February June 2018

Cheng Soon Ong & Christian Walder. Canberra February June 2018 Cheng Soon Ong & Christian Walder Research Group and College of Engineering and Computer Science Canberra February June 2018 Outlines Overview Introduction Linear Algebra Probability Linear Regression

More information

Chapter 9. Support Vector Machine. Yongdai Kim Seoul National University

Chapter 9. Support Vector Machine. Yongdai Kim Seoul National University Chapter 9. Support Vector Machine Yongdai Kim Seoul National University 1. Introduction Support Vector Machine (SVM) is a classification method developed by Vapnik (1996). It is thought that SVM improved

More information

Introduction to Support Vector Machines

Introduction to Support Vector Machines Introduction to Support Vector Machines Andreas Maletti Technische Universität Dresden Fakultät Informatik June 15, 2006 1 The Problem 2 The Basics 3 The Proposed Solution Learning by Machines Learning

More information

Analysis of Multiclass Support Vector Machines

Analysis of Multiclass Support Vector Machines Analysis of Multiclass Support Vector Machines Shigeo Abe Graduate School of Science and Technology Kobe University Kobe, Japan abe@eedept.kobe-u.ac.jp Abstract Since support vector machines for pattern

More information

Discriminative Direction for Kernel Classifiers

Discriminative Direction for Kernel Classifiers Discriminative Direction for Kernel Classifiers Polina Golland Artificial Intelligence Lab Massachusetts Institute of Technology Cambridge, MA 02139 polina@ai.mit.edu Abstract In many scientific and engineering

More information

Support Vector Machine for Classification and Regression

Support Vector Machine for Classification and Regression Support Vector Machine for Classification and Regression Ahlame Douzal AMA-LIG, Université Joseph Fourier Master 2R - MOSIG (2013) November 25, 2013 Loss function, Separating Hyperplanes, Canonical Hyperplan

More information

Incorporating Invariances in Nonlinear Support Vector Machines

Incorporating Invariances in Nonlinear Support Vector Machines Incorporating Invariances in Nonlinear Support Vector Machines Olivier Chapelle olivier.chapelle@lip6.fr LIP6, Paris, France Biowulf Technologies Bernhard Scholkopf bernhard.schoelkopf@tuebingen.mpg.de

More information

below, kernel PCA Eigenvectors, and linear combinations thereof. For the cases where the pre-image does exist, we can provide a means of constructing

below, kernel PCA Eigenvectors, and linear combinations thereof. For the cases where the pre-image does exist, we can provide a means of constructing Kernel PCA Pattern Reconstruction via Approximate Pre-Images Bernhard Scholkopf, Sebastian Mika, Alex Smola, Gunnar Ratsch, & Klaus-Robert Muller GMD FIRST, Rudower Chaussee 5, 12489 Berlin, Germany fbs,

More information

Support Vector Machines. Maximizing the Margin

Support Vector Machines. Maximizing the Margin Support Vector Machines Support vector achines (SVMs) learn a hypothesis: h(x) = b + Σ i= y i α i k(x, x i ) (x, y ),..., (x, y ) are the training exs., y i {, } b is the bias weight. α,..., α are the

More information

Support Vector Machine & Its Applications

Support Vector Machine & Its Applications Support Vector Machine & Its Applications A portion (1/3) of the slides are taken from Prof. Andrew Moore s SVM tutorial at http://www.cs.cmu.edu/~awm/tutorials Mingyue Tan The University of British Columbia

More information

Kernel Machines and Additive Fuzzy Systems: Classification and Function Approximation

Kernel Machines and Additive Fuzzy Systems: Classification and Function Approximation Kernel Machines and Additive Fuzzy Systems: Classification and Function Approximation Yixin Chen * and James Z. Wang * Dept. of Computer Science and Engineering, School of Information Sciences and Technology

More information

Constrained Optimization and Support Vector Machines

Constrained Optimization and Support Vector Machines Constrained Optimization and Support Vector Machines Man-Wai MAK Dept. of Electronic and Information Engineering, The Hong Kong Polytechnic University enmwmak@polyu.edu.hk http://www.eie.polyu.edu.hk/

More information

Support'Vector'Machines. Machine(Learning(Spring(2018 March(5(2018 Kasthuri Kannan

Support'Vector'Machines. Machine(Learning(Spring(2018 March(5(2018 Kasthuri Kannan Support'Vector'Machines Machine(Learning(Spring(2018 March(5(2018 Kasthuri Kannan kasthuri.kannan@nyumc.org Overview Support Vector Machines for Classification Linear Discrimination Nonlinear Discrimination

More information

Learning with kernels and SVM

Learning with kernels and SVM Learning with kernels and SVM Šámalova chata, 23. května, 2006 Petra Kudová Outline Introduction Binary classification Learning with Kernels Support Vector Machines Demo Conclusion Learning from data find

More information

Linear Dependency Between and the Input Noise in -Support Vector Regression

Linear Dependency Between and the Input Noise in -Support Vector Regression 544 IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 14, NO. 3, MAY 2003 Linear Dependency Between the Input Noise in -Support Vector Regression James T. Kwok Ivor W. Tsang Abstract In using the -support vector

More information

Neural Networks. Prof. Dr. Rudolf Kruse. Computational Intelligence Group Faculty for Computer Science

Neural Networks. Prof. Dr. Rudolf Kruse. Computational Intelligence Group Faculty for Computer Science Neural Networks Prof. Dr. Rudolf Kruse Computational Intelligence Group Faculty for Computer Science kruse@iws.cs.uni-magdeburg.de Rudolf Kruse Neural Networks 1 Supervised Learning / Support Vector Machines

More information

Support Vector Machines.

Support Vector Machines. Support Vector Machines www.cs.wisc.edu/~dpage 1 Goals for the lecture you should understand the following concepts the margin slack variables the linear support vector machine nonlinear SVMs the kernel

More information

Chemometrics: Classification of spectra

Chemometrics: Classification of spectra Chemometrics: Classification of spectra Vladimir Bochko Jarmo Alander University of Vaasa November 1, 2010 Vladimir Bochko Chemometrics: Classification 1/36 Contents Terminology Introduction Big picture

More information

Regularization and statistical learning theory for data analysis

Regularization and statistical learning theory for data analysis Computational Statistics & Data Analysis 38 (2002) 421 432 www.elsevier.com/locate/csda Regularization and statistical learning theory for data analysis Theodoros Evgeniou a;, Tomaso Poggio b, Massimiliano

More information

SVM TRADE-OFF BETWEEN MAXIMIZE THE MARGIN AND MINIMIZE THE VARIABLES USED FOR REGRESSION

SVM TRADE-OFF BETWEEN MAXIMIZE THE MARGIN AND MINIMIZE THE VARIABLES USED FOR REGRESSION International Journal of Pure and Applied Mathematics Volume 87 No. 6 2013, 741-750 ISSN: 1311-8080 (printed version); ISSN: 1314-3395 (on-line version) url: http://www.ijpam.eu doi: http://dx.doi.org/10.12732/ijpam.v87i6.2

More information

Hilbert Space Methods in Learning

Hilbert Space Methods in Learning Hilbert Space Methods in Learning guest lecturer: Risi Kondor 6772 Advanced Machine Learning and Perception (Jebara), Columbia University, October 15, 2003. 1 1. A general formulation of the learning problem

More information

Support Vector Machines

Support Vector Machines Wien, June, 2010 Paul Hofmarcher, Stefan Theussl, WU Wien Hofmarcher/Theussl SVM 1/21 Linear Separable Separating Hyperplanes Non-Linear Separable Soft-Margin Hyperplanes Hofmarcher/Theussl SVM 2/21 (SVM)

More information

Evaluation of simple performance measures for tuning SVM hyperparameters

Evaluation of simple performance measures for tuning SVM hyperparameters Neurocomputing 51 (2003) 41 59 www.elsevier.com/locate/neucom Evaluation of simple performance measures for tuning SVM hyperparameters Kaibo Duan, S. Sathiya Keerthi, Aun Neow Poo Department of Mechanical

More information

Support Vector Machines and Kernel Methods

Support Vector Machines and Kernel Methods 2018 CS420 Machine Learning, Lecture 3 Hangout from Prof. Andrew Ng. http://cs229.stanford.edu/notes/cs229-notes3.pdf Support Vector Machines and Kernel Methods Weinan Zhang Shanghai Jiao Tong University

More information

SUPPORT VECTOR MACHINE

SUPPORT VECTOR MACHINE SUPPORT VECTOR MACHINE Mainly based on https://nlp.stanford.edu/ir-book/pdf/15svm.pdf 1 Overview SVM is a huge topic Integration of MMDS, IIR, and Andrew Moore s slides here Our foci: Geometric intuition

More information

Learning with Rigorous Support Vector Machines

Learning with Rigorous Support Vector Machines Learning with Rigorous Support Vector Machines Jinbo Bi 1 and Vladimir N. Vapnik 2 1 Department of Mathematical Sciences, Rensselaer Polytechnic Institute, Troy NY 12180, USA 2 NEC Labs America, Inc. Princeton

More information

Support Vector Machines and Kernel Algorithms

Support Vector Machines and Kernel Algorithms Support Vector Machines and Kernel Algorithms Bernhard Schölkopf Max-Planck-Institut für biologische Kybernetik 72076 Tübingen, Germany Bernhard.Schoelkopf@tuebingen.mpg.de Alex Smola RSISE, Australian

More information

Lecture Support Vector Machine (SVM) Classifiers

Lecture Support Vector Machine (SVM) Classifiers Introduction to Machine Learning Lecturer: Amir Globerson Lecture 6 Fall Semester Scribe: Yishay Mansour 6.1 Support Vector Machine (SVM) Classifiers Classification is one of the most important tasks in

More information

Support Vector Machines (SVM) in bioinformatics. Day 1: Introduction to SVM

Support Vector Machines (SVM) in bioinformatics. Day 1: Introduction to SVM 1 Support Vector Machines (SVM) in bioinformatics Day 1: Introduction to SVM Jean-Philippe Vert Bioinformatics Center, Kyoto University, Japan Jean-Philippe.Vert@mines.org Human Genome Center, University

More information

Machine Learning Lecture 7

Machine Learning Lecture 7 Course Outline Machine Learning Lecture 7 Fundamentals (2 weeks) Bayes Decision Theory Probability Density Estimation Statistical Learning Theory 23.05.2016 Discriminative Approaches (5 weeks) Linear Discriminant

More information

Support Vector Machines for Classification: A Statistical Portrait

Support Vector Machines for Classification: A Statistical Portrait Support Vector Machines for Classification: A Statistical Portrait Yoonkyung Lee Department of Statistics The Ohio State University May 27, 2011 The Spring Conference of Korean Statistical Society KAIST,

More information

Performance Analysis of Spread Spectrum CDMA systems

Performance Analysis of Spread Spectrum CDMA systems 1 Performance Analysis of Spread Spectrum CDMA systems 16:33:546 Wireless Communication Technologies Spring 5 Instructor: Dr. Narayan Mandayam Summary by Liang Xiao lxiao@winlab.rutgers.edu WINLAB, Department

More information

Support Vector Machine. Natural Language Processing Lab lizhonghua

Support Vector Machine. Natural Language Processing Lab lizhonghua Support Vector Machine Natural Language Processing Lab lizhonghua Support Vector Machine Introduction Theory SVM primal and dual problem Parameter selection and practical issues Compare to other classifier

More information

Classifier Complexity and Support Vector Classifiers

Classifier Complexity and Support Vector Classifiers Classifier Complexity and Support Vector Classifiers Feature 2 6 4 2 0 2 4 6 8 RBF kernel 10 10 8 6 4 2 0 2 4 6 Feature 1 David M.J. Tax Pattern Recognition Laboratory Delft University of Technology D.M.J.Tax@tudelft.nl

More information

In this chapter, a support vector regression (SVR) based detector is presented for

In this chapter, a support vector regression (SVR) based detector is presented for Chapter 5 Support Vector Regression Approach to Large-MIMO Detection In this chapter, a support vector regression (SVR) based detector is presented for detection in large-mimo systems. The main motivation

More information

An Improved Conjugate Gradient Scheme to the Solution of Least Squares SVM

An Improved Conjugate Gradient Scheme to the Solution of Least Squares SVM An Improved Conjugate Gradient Scheme to the Solution of Least Squares SVM Wei Chu Chong Jin Ong chuwei@gatsby.ucl.ac.uk mpeongcj@nus.edu.sg S. Sathiya Keerthi mpessk@nus.edu.sg Control Division, Department

More information

FIND A FUNCTION TO CLASSIFY HIGH VALUE CUSTOMERS

FIND A FUNCTION TO CLASSIFY HIGH VALUE CUSTOMERS LINEAR CLASSIFIER 1 FIND A FUNCTION TO CLASSIFY HIGH VALUE CUSTOMERS x f y High Value Customers Salary Task: Find Nb Orders 150 70 300 100 200 80 120 100 Low Value Customers Salary Nb Orders 40 80 220

More information

Linear Classification and SVM. Dr. Xin Zhang

Linear Classification and SVM. Dr. Xin Zhang Linear Classification and SVM Dr. Xin Zhang Email: eexinzhang@scut.edu.cn What is linear classification? Classification is intrinsically non-linear It puts non-identical things in the same class, so a

More information

A Novel Blind Multiuser Detection Algorithm Based on Support Vector Machines

A Novel Blind Multiuser Detection Algorithm Based on Support Vector Machines A Novel Blind Multiuser Detection Algorithm Based on Support Vector Machines Yan Zhu, Limin Xiao, Shidong Zhou and Jing Wang Department of Electronic Engineering Tsinghua University Beijing, China 100084

More information

Vectors in many applications, most of the i, which are found by solving a quadratic program, turn out to be 0. Excellent classication accuracies in bo

Vectors in many applications, most of the i, which are found by solving a quadratic program, turn out to be 0. Excellent classication accuracies in bo Fast Approximation of Support Vector Kernel Expansions, and an Interpretation of Clustering as Approximation in Feature Spaces Bernhard Scholkopf 1, Phil Knirsch 2, Alex Smola 1, and Chris Burges 2 1 GMD

More information

ML (cont.): SUPPORT VECTOR MACHINES

ML (cont.): SUPPORT VECTOR MACHINES ML (cont.): SUPPORT VECTOR MACHINES CS540 Bryan R Gibson University of Wisconsin-Madison Slides adapted from those used by Prof. Jerry Zhu, CS540-1 1 / 40 Support Vector Machines (SVMs) The No-Math Version

More information

trailer to a given position [7]. This is a complicated classication problem (steering wheel left or right) when expressed in cartesian coordinates in

trailer to a given position [7]. This is a complicated classication problem (steering wheel left or right) when expressed in cartesian coordinates in Support Vector Methods in Learning and Feature Extraction Bernhard Scholkopf x y, Alex Smola x y, Klaus-Robert Muller x, Chris Burges z, Vladimir Vapnik? x GMD FIRST, Rudower Chaussee 5, 12489 Berlin,

More information

Robust Kernel-Based Regression

Robust Kernel-Based Regression Robust Kernel-Based Regression Budi Santosa Department of Industrial Engineering Sepuluh Nopember Institute of Technology Kampus ITS Surabaya Surabaya 60111,Indonesia Theodore B. Trafalis School of Industrial

More information

Support Vector Machines for Classification and Regression

Support Vector Machines for Classification and Regression CIS 520: Machine Learning Oct 04, 207 Support Vector Machines for Classification and Regression Lecturer: Shivani Agarwal Disclaimer: These notes are designed to be a supplement to the lecture. They may

More information

Discussion of Some Problems About Nonlinear Time Series Prediction Using ν-support Vector Machine

Discussion of Some Problems About Nonlinear Time Series Prediction Using ν-support Vector Machine Commun. Theor. Phys. (Beijing, China) 48 (2007) pp. 117 124 c International Academic Publishers Vol. 48, No. 1, July 15, 2007 Discussion of Some Problems About Nonlinear Time Series Prediction Using ν-support

More information

Linear Regression and Its Applications

Linear Regression and Its Applications Linear Regression and Its Applications Predrag Radivojac October 13, 2014 Given a data set D = {(x i, y i )} n the objective is to learn the relationship between features and the target. We usually start

More information

Brief Introduction to Machine Learning

Brief Introduction to Machine Learning Brief Introduction to Machine Learning Yuh-Jye Lee Lab of Data Science and Machine Intelligence Dept. of Applied Math. at NCTU August 29, 2016 1 / 49 1 Introduction 2 Binary Classification 3 Support Vector

More information

Discriminative Models

Discriminative Models No.5 Discriminative Models Hui Jiang Department of Electrical Engineering and Computer Science Lassonde School of Engineering York University, Toronto, Canada Outline Generative vs. Discriminative models

More information

Support Vector Machines Explained

Support Vector Machines Explained December 23, 2008 Support Vector Machines Explained Tristan Fletcher www.cs.ucl.ac.uk/staff/t.fletcher/ Introduction This document has been written in an attempt to make the Support Vector Machines (SVM),

More information

Discriminative Models

Discriminative Models No.5 Discriminative Models Hui Jiang Department of Electrical Engineering and Computer Science Lassonde School of Engineering York University, Toronto, Canada Outline Generative vs. Discriminative models

More information

Multi User Detection I

Multi User Detection I January 12, 2005 Outline Overview Multiple Access Communication Motivation: What is MU Detection? Overview of DS/CDMA systems Concept and Codes used in CDMA CDMA Channels Models Synchronous and Asynchronous

More information

Support Vector Machines

Support Vector Machines Support Vector Machines Statistical Inference with Reproducing Kernel Hilbert Space Kenji Fukumizu Institute of Statistical Mathematics, ROIS Department of Statistical Science, Graduate University for

More information

Relevance Vector Machines for Earthquake Response Spectra

Relevance Vector Machines for Earthquake Response Spectra 2012 2011 American American Transactions Transactions on on Engineering Engineering & Applied Applied Sciences Sciences. American Transactions on Engineering & Applied Sciences http://tuengr.com/ateas

More information

On the V γ Dimension for Regression in Reproducing Kernel Hilbert Spaces. Theodoros Evgeniou, Massimiliano Pontil

On the V γ Dimension for Regression in Reproducing Kernel Hilbert Spaces. Theodoros Evgeniou, Massimiliano Pontil MASSACHUSETTS INSTITUTE OF TECHNOLOGY ARTIFICIAL INTELLIGENCE LABORATORY and CENTER FOR BIOLOGICAL AND COMPUTATIONAL LEARNING DEPARTMENT OF BRAIN AND COGNITIVE SCIENCES A.I. Memo No. 1656 May 1999 C.B.C.L

More information

The Decision List Machine

The Decision List Machine The Decision List Machine Marina Sokolova SITE, University of Ottawa Ottawa, Ont. Canada,K1N-6N5 sokolova@site.uottawa.ca Nathalie Japkowicz SITE, University of Ottawa Ottawa, Ont. Canada,K1N-6N5 nat@site.uottawa.ca

More information

Content. Learning. Regression vs Classification. Regression a.k.a. function approximation and Classification a.k.a. pattern recognition

Content. Learning. Regression vs Classification. Regression a.k.a. function approximation and Classification a.k.a. pattern recognition Content Andrew Kusiak Intelligent Systems Laboratory 239 Seamans Center The University of Iowa Iowa City, IA 52242-527 andrew-kusiak@uiowa.edu http://www.icaen.uiowa.edu/~ankusiak Introduction to learning

More information

ANALYSIS OF A PARTIAL DECORRELATOR IN A MULTI-CELL DS/CDMA SYSTEM

ANALYSIS OF A PARTIAL DECORRELATOR IN A MULTI-CELL DS/CDMA SYSTEM ANAYSIS OF A PARTIA DECORREATOR IN A MUTI-CE DS/CDMA SYSTEM Mohammad Saquib ECE Department, SU Baton Rouge, A 70803-590 e-mail: saquib@winlab.rutgers.edu Roy Yates WINAB, Rutgers University Piscataway

More information

Lecture 9: Large Margin Classifiers. Linear Support Vector Machines

Lecture 9: Large Margin Classifiers. Linear Support Vector Machines Lecture 9: Large Margin Classifiers. Linear Support Vector Machines Perceptrons Definition Perceptron learning rule Convergence Margin & max margin classifiers (Linear) support vector machines Formulation

More information

L5 Support Vector Classification

L5 Support Vector Classification L5 Support Vector Classification Support Vector Machine Problem definition Geometrical picture Optimization problem Optimization Problem Hard margin Convexity Dual problem Soft margin problem Alexander

More information

NON-FIXED AND ASYMMETRICAL MARGIN APPROACH TO STOCK MARKET PREDICTION USING SUPPORT VECTOR REGRESSION. Haiqin Yang, Irwin King and Laiwan Chan

NON-FIXED AND ASYMMETRICAL MARGIN APPROACH TO STOCK MARKET PREDICTION USING SUPPORT VECTOR REGRESSION. Haiqin Yang, Irwin King and Laiwan Chan In The Proceedings of ICONIP 2002, Singapore, 2002. NON-FIXED AND ASYMMETRICAL MARGIN APPROACH TO STOCK MARKET PREDICTION USING SUPPORT VECTOR REGRESSION Haiqin Yang, Irwin King and Laiwan Chan Department

More information

Linear vs Non-linear classifier. CS789: Machine Learning and Neural Network. Introduction

Linear vs Non-linear classifier. CS789: Machine Learning and Neural Network. Introduction Linear vs Non-linear classifier CS789: Machine Learning and Neural Network Support Vector Machine Jakramate Bootkrajang Department of Computer Science Chiang Mai University Linear classifier is in the

More information

Linear classifiers selecting hyperplane maximizing separation margin between classes (large margin classifiers)

Linear classifiers selecting hyperplane maximizing separation margin between classes (large margin classifiers) Support vector machines In a nutshell Linear classifiers selecting hyperplane maximizing separation margin between classes (large margin classifiers) Solution only depends on a small subset of training

More information

Support Vector Machines for Classification and Regression. 1 Linearly Separable Data: Hard Margin SVMs

Support Vector Machines for Classification and Regression. 1 Linearly Separable Data: Hard Margin SVMs E0 270 Machine Learning Lecture 5 (Jan 22, 203) Support Vector Machines for Classification and Regression Lecturer: Shivani Agarwal Disclaimer: These notes are a brief summary of the topics covered in

More information

Improvements to Platt s SMO Algorithm for SVM Classifier Design

Improvements to Platt s SMO Algorithm for SVM Classifier Design LETTER Communicated by John Platt Improvements to Platt s SMO Algorithm for SVM Classifier Design S. S. Keerthi Department of Mechanical and Production Engineering, National University of Singapore, Singapore-119260

More information

Learning Kernel Parameters by using Class Separability Measure

Learning Kernel Parameters by using Class Separability Measure Learning Kernel Parameters by using Class Separability Measure Lei Wang, Kap Luk Chan School of Electrical and Electronic Engineering Nanyang Technological University Singapore, 3979 E-mail: P 3733@ntu.edu.sg,eklchan@ntu.edu.sg

More information

Solving the SVM Optimization Problem

Solving the SVM Optimization Problem Solving the SVM Optimization Problem Kernel-based Learning Methods Christian Igel Institut für Neuroinformatik Ruhr-Universität Bochum, Germany http://www.neuroinformatik.rub.de July 16, 2009 Christian

More information

MinOver Revisited for Incremental Support-Vector-Classification

MinOver Revisited for Incremental Support-Vector-Classification MinOver Revisited for Incremental Support-Vector-Classification Thomas Martinetz Institute for Neuro- and Bioinformatics University of Lübeck D-23538 Lübeck, Germany martinetz@informatik.uni-luebeck.de

More information

Machine Learning 4771

Machine Learning 4771 Machine Learning 477 Instructor: Tony Jebara Topic 5 Generalization Guarantees VC-Dimension Nearest Neighbor Classification (infinite VC dimension) Structural Risk Minimization Support Vector Machines

More information

Sparse Support Vector Machines by Kernel Discriminant Analysis

Sparse Support Vector Machines by Kernel Discriminant Analysis Sparse Support Vector Machines by Kernel Discriminant Analysis Kazuki Iwamura and Shigeo Abe Kobe University - Graduate School of Engineering Kobe, Japan Abstract. We discuss sparse support vector machines

More information

CIS 520: Machine Learning Oct 09, Kernel Methods

CIS 520: Machine Learning Oct 09, Kernel Methods CIS 520: Machine Learning Oct 09, 207 Kernel Methods Lecturer: Shivani Agarwal Disclaimer: These notes are designed to be a supplement to the lecture They may or may not cover all the material discussed

More information

Support Vector Machines

Support Vector Machines EE 17/7AT: Optimization Models in Engineering Section 11/1 - April 014 Support Vector Machines Lecturer: Arturo Fernandez Scribe: Arturo Fernandez 1 Support Vector Machines Revisited 1.1 Strictly) Separable

More information

Support Vector Machine I

Support Vector Machine I Support Vector Machine I Statistical Data Analysis with Positive Definite Kernels Kenji Fukumizu Institute of Statistical Mathematics, ROIS Department of Statistical Science, Graduate University for Advanced

More information

Ramp Loss SVM with L1-Norm Regularizaion

Ramp Loss SVM with L1-Norm Regularizaion Virginia Commonwealth University VCU Scholars Compass Theses and Dissertations Graduate School 2014 Ramp Loss SVM with L1-Norm Regularizaion Eric Hess Virginia Commonwealth University Follow this and additional

More information

Support Vector Machines with Example Dependent Costs

Support Vector Machines with Example Dependent Costs Support Vector Machines with Example Dependent Costs Ulf Brefeld, Peter Geibel, and Fritz Wysotzki TU Berlin, Fak. IV, ISTI, AI Group, Sekr. FR5-8 Franklinstr. 8/9, D-0587 Berlin, Germany Email {geibel

More information

Outliers Treatment in Support Vector Regression for Financial Time Series Prediction

Outliers Treatment in Support Vector Regression for Financial Time Series Prediction Outliers Treatment in Support Vector Regression for Financial Time Series Prediction Haiqin Yang, Kaizhu Huang, Laiwan Chan, Irwin King, and Michael R. Lyu Department of Computer Science and Engineering

More information

Support Vector Machines. Machine Learning Fall 2017

Support Vector Machines. Machine Learning Fall 2017 Support Vector Machines Machine Learning Fall 2017 1 Where are we? Learning algorithms Decision Trees Perceptron AdaBoost 2 Where are we? Learning algorithms Decision Trees Perceptron AdaBoost Produce

More information

LINEAR CLASSIFIERS. J. Elder CSE 4404/5327 Introduction to Machine Learning and Pattern Recognition

LINEAR CLASSIFIERS. J. Elder CSE 4404/5327 Introduction to Machine Learning and Pattern Recognition LINEAR CLASSIFIERS Classification: Problem Statement 2 In regression, we are modeling the relationship between a continuous input variable x and a continuous target variable t. In classification, the input

More information

Introduction to SVM and RVM

Introduction to SVM and RVM Introduction to SVM and RVM Machine Learning Seminar HUS HVL UIB Yushu Li, UIB Overview Support vector machine SVM First introduced by Vapnik, et al. 1992 Several literature and wide applications Relevance

More information

SUPPORT VECTOR MACHINE FOR THE SIMULTANEOUS APPROXIMATION OF A FUNCTION AND ITS DERIVATIVE

SUPPORT VECTOR MACHINE FOR THE SIMULTANEOUS APPROXIMATION OF A FUNCTION AND ITS DERIVATIVE SUPPORT VECTOR MACHINE FOR THE SIMULTANEOUS APPROXIMATION OF A FUNCTION AND ITS DERIVATIVE M. Lázaro 1, I. Santamaría 2, F. Pérez-Cruz 1, A. Artés-Rodríguez 1 1 Departamento de Teoría de la Señal y Comunicaciones

More information

Iteratively Reweighted Least Square for Asymmetric L 2 -Loss Support Vector Regression

Iteratively Reweighted Least Square for Asymmetric L 2 -Loss Support Vector Regression Iteratively Reweighted Least Square for Asymmetric L 2 -Loss Support Vector Regression Songfeng Zheng Department of Mathematics Missouri State University Springfield, MO 65897 SongfengZheng@MissouriState.edu

More information

Model Selection for LS-SVM : Application to Handwriting Recognition

Model Selection for LS-SVM : Application to Handwriting Recognition Model Selection for LS-SVM : Application to Handwriting Recognition Mathias M. Adankon and Mohamed Cheriet Synchromedia Laboratory for Multimedia Communication in Telepresence, École de Technologie Supérieure,

More information

Support Vector Machine (continued)

Support Vector Machine (continued) Support Vector Machine continued) Overlapping class distribution: In practice the class-conditional distributions may overlap, so that the training data points are no longer linearly separable. We need

More information

Introduction to Support Vector Machines

Introduction to Support Vector Machines Introduction to Support Vector Machines Hsuan-Tien Lin Learning Systems Group, California Institute of Technology Talk in NTU EE/CS Speech Lab, November 16, 2005 H.-T. Lin (Learning Systems Group) Introduction

More information

Lecture 10: Support Vector Machine and Large Margin Classifier

Lecture 10: Support Vector Machine and Large Margin Classifier Lecture 10: Support Vector Machine and Large Margin Classifier Applied Multivariate Analysis Math 570, Fall 2014 Xingye Qiao Department of Mathematical Sciences Binghamton University E-mail: qiao@math.binghamton.edu

More information

This is an author-deposited version published in : Eprints ID : 17710

This is an author-deposited version published in :   Eprints ID : 17710 Open Archive TOULOUSE Archive Ouverte (OATAO) OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible. This is an author-deposited

More information