Numerical Analysis

· Course Outline

· *Introduction pdf *pdf

· *Conjugate Gradient Method for Solving Linear System pdf *pdf pdf code Note_EX1

· IEEE Xplore - Training feedforward networks with the Marquardt algorithm *Deep Learning Toolbox

· Gradient of deep perceptrons

· Gradients in Neural mapping *my_net2.m *perceptrons.m

· Nonlinear System Solving by Newton's Method *pdf Homework 2 nonlinear_newton

· Linear mapping and nonlinear mapping

· object-oriented design of deep learning

· Least mean square method of deep neural networks

· Batch Updating and Autoencoding

· My_Net_Train_Test

· Final - Neural Network Toolbox & Matconvnet

· * Matlab_programing pdf *pdf Homework 1

· *Line & hyperplane fitting LDA for wine classification 2016fit_tanh_M Project2016

o demo_hp_fitting.m

o demo_hp_fitting2.m

o demo_wine_fitting.m

o plot2d.m

o MLP_Tool.m MLP_Tool.fig

o Wine.dat

· Function approximation using NNSYSID

o The NNSYSID Toolbox

o eval_MLP2

o mean_square_error2.m

o learn_myMLP.m

· *Radial Basis Function Network - File Exchange - MATLAB Central

· Gradient Descent method and nonlinear conjugate gradient method for learning RBF pdf

· Nonlinear_Conjugate_Gradient_Method (pdf)

· Implementation_of_NCG_method (pdf)

· Midterm_2012

· Nonlinear Conjugate Gradient Learning

· *LM learning of MLP and RBF neural networks Homework 3

o MATLAB coding (LMMLP code) LMMLP code2014 My LM MLP My LM RBF

1. GA optimization of RBF

2. LM-learning of MLP pdf Evaluation

3. LM-learning of RBF

o Exercise

· DA using NNSYSID Separable DA_LM_RBF

· Nonlinear Recursive Relations Hill Valley_LM RBF *K-means *Advanced clustering data_gen.m annealed_kmeans2.m color_clustering.rar application 期末報告

· *MFA(mean field annealing) optimization pdf

o MFA optimization for Graph Bisection pdf

o Matlab coding for graph bisection pdf

o MFA Optimization for TSP pdf

· *EM clustering pdf

· Unsupervised MFA learning for data clustering pdf

o *Learning generative models *Annealed Kullback-Leibler divergence minimization

o plot_normal.m

o GM.m

o microarrayPT.mat

o demo_kmeans

o Exercise

· Unsupervised MFA learning for self-organization and ICA & supervised MFA learning for Classification pdf

· Unsupervised MFA learning for self-organization

· Supervised MFA learning for classification

· Unsupervised MFA learning for independent component analysis

· Final project: LM learning for RBF networks

· Matlab programming of RBF-LM learning

· Data driven function approximation

· Reference

· Slides for MATLAB programming I

· Slides for MATLAB programming II

· Slides for MATLAB programming II

· Numerical mathematics and computing, four Edition, Cheney & Kincaid

· Online slides

· Lecture slides: http://134.208.26.61/course/na/2007NA.htm

· Numerical methods with MATLAB:Implementation and application, G. Rechtenwald

· http://web.cecs.pdx.edu/~gerry/nmm/

· Numerical Methods with MATLAB

Course Outline

科目代碼	AM__52100
科目名稱	數值分析
授課老師	吳建銘
開課班級	碩士
每週授課時數	3(二8/二9/二10)
校內分機	3531
教師電子郵件	jmwu@livemail. tw
教師辦公室	理B301
會談時間
課程助教	吳俊樟
助教電子郵件
助教工作項目
課程目標	This course focuses on fundamental numerical methods and advanced iterative approaches for intelligence numerical computation. It will first introduce numerical and symbolic integration, differentiation and iterative approaches respectively for solving hyper-plane fitting, linear systems and nonlinear systems, further addressing on nonlinear function approximation, unconstrained optimizations based on the gradient method, the Newton-Gauss method and the Levenberg-Marquardt method and annealed Kullback-Leibler divergence minimization for solving data clustering, density estimation, independent component analysis, computational self-organization and classification.
教學方法	Lectures and matlab programming exercises
教學評量	Midterm 30%+Final 30%+ Exercise 30%+ General Performance 10%
課堂教材	Textbook and lecture slides
其他教材	數值分析
作業備註
其他標題
其他內容

1		進度	重要事項
1	1	Introduction
2	2	Matlab programming	Exercise 1
3	3	Nonlinear System Solving
4	4	Line fitting, hyperplane fitting, quadratic-surface fitting	Exercise 2
5	5	Function approximation
6	6	Newton method and Newton-Gauss method	Exercise 3
7	7	Levenberg-Marquardt method
8	8	Recursive function approximation	Exercise 4
9	9	期中考試週 Midterm Exam
10	10	Differential equation approximation
11	11	Classification I	Exercise 5
12	12	Classifiaction II
13	13	Kullback-Leibler divergence minimization	Exercise 6
14	14	Mean field annealing
15	15	Clsutering and density function approximation	Exercise 7
16	16	Self-organization
17	17	Applications	Exercise 8
	18	期末考試週