Lecture 25 Radial Basis Network (II)

Slides:



Advertisements
Similar presentations
Support Vector Machines
Advertisements

Graph Laplacian Regularization for Large-Scale Semidefinite Programming Kilian Weinberger et al. NIPS 2006 presented by Aggeliki Tsoli.
RBF Neural Networks x x1 Examples inside circles 1 and 2 are of class +, examples outside both circles are of class – What NN does.
Radial Basis Functions
Giansalvo EXIN Cirrincione unit #6 Problem: given a mapping: x  d  t  and a TS of N points, find a function h(x) such that: h(x n ) = t n n = 1,
November 2, 2010Neural Networks Lecture 14: Radial Basis Functions 1 Cascade Correlation Weights to each new hidden node are trained to maximize the covariance.
Hazırlayan NEURAL NETWORKS Radial Basis Function Networks I PROF. DR. YUSUF OYSAL.
Intro. ANN & Fuzzy Systems Lecture 21 Clustering (2)
Aula 4 Radial Basis Function Networks
Radial Basis Function (RBF) Networks
Radial-Basis Function Networks
Hazırlayan NEURAL NETWORKS Radial Basis Function Networks II PROF. DR. YUSUF OYSAL.
Radial Basis Function Networks
8/10/ RBF NetworksM.W. Mak Radial Basis Function Networks 1. Introduction 2. Finding RBF Parameters 3. Decision Surface of RBF Networks 4. Comparison.
Today Wrap up of probability Vectors, Matrices. Calculus

Radial Basis Function Networks
Artificial Neural Networks Shreekanth Mandayam Robi Polikar …… …... … net k   
MML Inference of RBFs Enes Makalic Lloyd Allison Andrew Paplinski.
Least-Mean-Square Training of Cluster-Weighted-Modeling National Taiwan University Department of Computer Science and Information Engineering.
Intro. ANN & Fuzzy Systems Lecture 23 Clustering (4)
Intro. ANN & Fuzzy Systems Lecture 14. MLP (VI): Model Selection.
Lecture 2: Statistical learning primer for biologists
Computational Intelligence Winter Term 2015/16 Prof. Dr. Günter Rudolph Lehrstuhl für Algorithm Engineering (LS 11) Fakultät für Informatik TU Dortmund.
Introduction to Radial Basis Function Networks
Intelligent Numerical Computation1 Center:Width:.
Intro. ANN & Fuzzy Systems Lecture 38 Mixture of Experts Neural Network.
Intro. ANN & Fuzzy Systems Lecture 16. Classification (II): Practical Considerations.
Numerical Analysis – Data Fitting Hanyang University Jong-Il Park.
Intro. ANN & Fuzzy Systems Lecture 24 Radial Basis Network (I)
1 Nonlinear Sub-optimal Mid Course Guidance with Desired Alinement using MPQC P. N. Dwivedi, Dr. A.Bhattacharya, Scientist, DRDO, Hyderabad-,INDIA Dr.
Regularized Least-Squares and Convex Optimization.
Intro. ANN & Fuzzy Systems Lecture 11. MLP (III): Back-Propagation.
Pattern Recognition Lecture 20: Neural Networks 3 Dr. Richard Spillman Pacific Lutheran University.
CSE343/543 Machine Learning Mayank Vatsa Lecture slides are prepared using several teaching resources and no authorship is claimed for any slides.
Today’s Lecture Neural networks Training
Neural networks and support vector machines
Deep Feedforward Networks
Supervised Learning in ANNs
12. Principles of Parameter Estimation
Introduction to Radial Basis Function Networks
Neural Networks Winter-Spring 2014
Lecture 12. MLP (IV): Programming & Implementation
Lecture 12. MLP (IV): Programming & Implementation
Lecture 11. MLP (III): Back-Propagation
Disadvantages of Discrete Neurons
Computational Intelligence
Hyperparameters, bias-variance tradeoff, validation
Outline Single neuron case: Nonlinear error correcting learning
Lecture 24 Radial Basis Network (I)
Neuro-Computing Lecture 4 Radial Basis Function Network
10701 / Machine Learning Today: - Cross validation,
Neural Network - 2 Mayank Vatsa
Lecture 7. Learning (IV): Error correcting Learning and LMS Algorithm
Computational Intelligence
Numerical Analysis Lecture13.
Capabilities of Threshold Neurons
Chapter 8: Generalization and Function Approximation
Lecture 5. Learning (II) Sampling
Lecture 18. SVM (II): Non-separable Cases
Learning Theory Reza Shadmehr
Machine Learning: UNIT-4 CHAPTER-1
Introduction to Radial Basis Function Networks
Recursively Adapted Radial Basis Function Networks and its Relationship to Resource Allocating Networks and Online Kernel Learning Weifeng Liu, Puskal.
Computational Intelligence
Computer Vision Lecture 19: Object Recognition III
Mathematical Foundations of BME
12. Principles of Parameter Estimation
Lecture 16. Classification (II): Practical Considerations
Computational Intelligence
Presentation transcript:

Lecture 25 Radial Basis Network (II)

Outline Regularization Network Formulation Radial Basis Network Type 2 Generalized RBF network Training algorithm Implementation details (C) 2001 by Yu Hen Hu

Properties of Regularization network An RBF network is a universal approximator: it can approximate arbitrarily well any multivariate continuous function on a compact support in Rn where n is the dimension of feature vectors, given sufficient number of hidden neurons. It is optimal in that it minimizes E(F). It also has the best approximation property. That means given an unknown nonlinear function f, there always exists a choice of RBF coefficients that approximates f better than other possible choices of models. (C) 2001 by Yu Hen Hu

Radial Basis Network (Type II) Instead of xi, use virtual data points tj in the solution of F(x). Define Substitute each xi into eq. F(xi)=di we have a new system: (GTG + lGo)w = GTd Thus, w = (GTG + lGo)-1GTd when l = 0, w = G+d = (GTG)-1GTd where G+ is the pseudo-inverse matrix of G. (C) 2001 by Yu Hen Hu

RBN2 Algorithm Summary Given: {xi; 1  i  K}, d: desired output, and J: # of radial basis neurons Cluster {xi} into J clusters, find clustering centers {tj;1  j  J}. Variance j2 or inverse covariance matrix j1 are also computed. Compute G matrix (K by J) and G0 matrix. Gi,j+1 = exp(0.5||x(i)tj||2/j2) or Gi,j+1 = exp(0.5(x(i)tj)Tj1(x(i)tj)) Solve w = G†d or (GTG + lG0)-1GTd Above procedure can be refined by fitting the clusters into a Gaussian mixture model and train it with the EM algorithm. (C) 2001 by Yu Hen Hu

Example (C) 2001 by Yu Hen Hu

General RBF Network Consider a Gaussian RBF model In RBN-II training, in order to compute {wi}, parameters {tj} are determined in advance using Kmeans clustering and s2 is selected initially. To fit the model better at F(xi) = di, these parameters may need fine-tuning. Additional enhancements include Allowing each basis has its own width parameter sj, and A bias term is added to compensate for nonzero background value of the function over the support. While similar to the Gaussian mixture model, {wi} can be negative is the main difference. (C) 2001 by Yu Hen Hu

Training of Generalized RBN The parameters q = {wj, tj, sj, b} are to be chosen to minimize the approximation error The steepest descent gradient method leads to: Specifically, for 1  m  J (C) 2001 by Yu Hen Hu

Training … Note that Hence Thus, the individual parameters’ on-line learning formula are: (C) 2001 by Yu Hen Hu

Implementation Details The cost function may be augmented with additional smoothing terms for the purpose of regularization. For example, the derivative of F(x|q) may be bounded by a user-specified constant. However, this will make the training formula more complicated. Initialization of RBF centers and variance can be accomplished using the Kmeans clustering algorithm Selection of the number of RBF function is part of the regularization process and often need to be done using trail-and-error, or heuristics. Cross-validation may also be used to give a more objective criterion. A feasible range may be imposed on each parameter to prevent numerical problem. E.g. s2  e > 0 (C) 2001 by Yu Hen Hu