|
wolffd@0
|
1 % Netlab Toolbox
|
|
wolffd@0
|
2 % Version 3.3.1 18-Jun-2004
|
|
wolffd@0
|
3 %
|
|
wolffd@0
|
4 % conffig - Display a confusion matrix.
|
|
wolffd@0
|
5 % confmat - Compute a confusion matrix.
|
|
wolffd@0
|
6 % conjgrad - Conjugate gradients optimization.
|
|
wolffd@0
|
7 % consist - Check that arguments are consistent.
|
|
wolffd@0
|
8 % convertoldnet- Convert pre-2.3 release MLP and MDN nets to new format
|
|
wolffd@0
|
9 % datread - Read data from an ascii file.
|
|
wolffd@0
|
10 % datwrite - Write data to ascii file.
|
|
wolffd@0
|
11 % dem2ddat - Generates two dimensional data for demos.
|
|
wolffd@0
|
12 % demard - Automatic relevance determination using the MLP.
|
|
wolffd@0
|
13 % demev1 - Demonstrate Bayesian regression for the MLP.
|
|
wolffd@0
|
14 % demev2 - Demonstrate Bayesian classification for the MLP.
|
|
wolffd@0
|
15 % demev3 - Demonstrate Bayesian regression for the RBF.
|
|
wolffd@0
|
16 % demgauss - Demonstrate sampling from Gaussian distributions.
|
|
wolffd@0
|
17 % demglm1 - Demonstrate simple classification using a generalized linear model.
|
|
wolffd@0
|
18 % demglm2 - Demonstrate simple classification using a generalized linear model.
|
|
wolffd@0
|
19 % demgmm1 - Demonstrate density modelling with a Gaussian mixture model.
|
|
wolffd@0
|
20 % demgmm3 - Demonstrate density modelling with a Gaussian mixture model.
|
|
wolffd@0
|
21 % demgmm4 - Demonstrate density modelling with a Gaussian mixture model.
|
|
wolffd@0
|
22 % demgmm5 - Demonstrate density modelling with a PPCA mixture model.
|
|
wolffd@0
|
23 % demgp - Demonstrate simple regression using a Gaussian Process.
|
|
wolffd@0
|
24 % demgpard - Demonstrate ARD using a Gaussian Process.
|
|
wolffd@0
|
25 % demgpot - Computes the gradient of the negative log likelihood for a mixture model.
|
|
wolffd@0
|
26 % demgtm1 - Demonstrate EM for GTM.
|
|
wolffd@0
|
27 % demgtm2 - Demonstrate GTM for visualisation.
|
|
wolffd@0
|
28 % demhint - Demonstration of Hinton diagram for 2-layer feed-forward network.
|
|
wolffd@0
|
29 % demhmc1 - Demonstrate Hybrid Monte Carlo sampling on mixture of two Gaussians.
|
|
wolffd@0
|
30 % demhmc2 - Demonstrate Bayesian regression with Hybrid Monte Carlo sampling.
|
|
wolffd@0
|
31 % demhmc3 - Demonstrate Bayesian regression with Hybrid Monte Carlo sampling.
|
|
wolffd@0
|
32 % demkmean - Demonstrate simple clustering model trained with K-means.
|
|
wolffd@0
|
33 % demknn1 - Demonstrate nearest neighbour classifier.
|
|
wolffd@0
|
34 % demmdn1 - Demonstrate fitting a multi-valued function using a Mixture Density Network.
|
|
wolffd@0
|
35 % demmet1 - Demonstrate Markov Chain Monte Carlo sampling on a Gaussian.
|
|
wolffd@0
|
36 % demmlp1 - Demonstrate simple regression using a multi-layer perceptron
|
|
wolffd@0
|
37 % demmlp2 - Demonstrate simple classification using a multi-layer perceptron
|
|
wolffd@0
|
38 % demnlab - A front-end Graphical User Interface to the demos
|
|
wolffd@0
|
39 % demns1 - Demonstrate Neuroscale for visualisation.
|
|
wolffd@0
|
40 % demolgd1 - Demonstrate simple MLP optimisation with on-line gradient descent
|
|
wolffd@0
|
41 % demopt1 - Demonstrate different optimisers on Rosenbrock's function.
|
|
wolffd@0
|
42 % dempot - Computes the negative log likelihood for a mixture model.
|
|
wolffd@0
|
43 % demprgp - Demonstrate sampling from a Gaussian Process prior.
|
|
wolffd@0
|
44 % demprior - Demonstrate sampling from a multi-parameter Gaussian prior.
|
|
wolffd@0
|
45 % demrbf1 - Demonstrate simple regression using a radial basis function network.
|
|
wolffd@0
|
46 % demsom1 - Demonstrate SOM for visualisation.
|
|
wolffd@0
|
47 % demtrain - Demonstrate training of MLP network.
|
|
wolffd@0
|
48 % dist2 - Calculates squared distance between two sets of points.
|
|
wolffd@0
|
49 % eigdec - Sorted eigendecomposition
|
|
wolffd@0
|
50 % errbayes - Evaluate Bayesian error function for network.
|
|
wolffd@0
|
51 % evidence - Re-estimate hyperparameters using evidence approximation.
|
|
wolffd@0
|
52 % fevbayes - Evaluate Bayesian regularisation for network forward propagation.
|
|
wolffd@0
|
53 % gauss - Evaluate a Gaussian distribution.
|
|
wolffd@0
|
54 % gbayes - Evaluate gradient of Bayesian error function for network.
|
|
wolffd@0
|
55 % glm - Create a generalized linear model.
|
|
wolffd@0
|
56 % glmderiv - Evaluate derivatives of GLM outputs with respect to weights.
|
|
wolffd@0
|
57 % glmerr - Evaluate error function for generalized linear model.
|
|
wolffd@0
|
58 % glmevfwd - Forward propagation with evidence for GLM
|
|
wolffd@0
|
59 % glmfwd - Forward propagation through generalized linear model.
|
|
wolffd@0
|
60 % glmgrad - Evaluate gradient of error function for generalized linear model.
|
|
wolffd@0
|
61 % glmhess - Evaluate the Hessian matrix for a generalised linear model.
|
|
wolffd@0
|
62 % glminit - Initialise the weights in a generalized linear model.
|
|
wolffd@0
|
63 % glmpak - Combines weights and biases into one weights vector.
|
|
wolffd@0
|
64 % glmtrain - Specialised training of generalized linear model
|
|
wolffd@0
|
65 % glmunpak - Separates weights vector into weight and bias matrices.
|
|
wolffd@0
|
66 % gmm - Creates a Gaussian mixture model with specified architecture.
|
|
wolffd@0
|
67 % gmmactiv - Computes the activations of a Gaussian mixture model.
|
|
wolffd@0
|
68 % gmmem - EM algorithm for Gaussian mixture model.
|
|
wolffd@0
|
69 % gmminit - Initialises Gaussian mixture model from data
|
|
wolffd@0
|
70 % gmmpak - Combines all the parameters in a Gaussian mixture model into one vector.
|
|
wolffd@0
|
71 % gmmpost - Computes the class posterior probabilities of a Gaussian mixture model.
|
|
wolffd@0
|
72 % gmmprob - Computes the data probability for a Gaussian mixture model.
|
|
wolffd@0
|
73 % gmmsamp - Sample from a Gaussian mixture distribution.
|
|
wolffd@0
|
74 % gmmunpak - Separates a vector of Gaussian mixture model parameters into its components.
|
|
wolffd@0
|
75 % gp - Create a Gaussian Process.
|
|
wolffd@0
|
76 % gpcovar - Calculate the covariance for a Gaussian Process.
|
|
wolffd@0
|
77 % gpcovarf - Calculate the covariance function for a Gaussian Process.
|
|
wolffd@0
|
78 % gpcovarp - Calculate the prior covariance for a Gaussian Process.
|
|
wolffd@0
|
79 % gperr - Evaluate error function for Gaussian Process.
|
|
wolffd@0
|
80 % gpfwd - Forward propagation through Gaussian Process.
|
|
wolffd@0
|
81 % gpgrad - Evaluate error gradient for Gaussian Process.
|
|
wolffd@0
|
82 % gpinit - Initialise Gaussian Process model.
|
|
wolffd@0
|
83 % gppak - Combines GP hyperparameters into one vector.
|
|
wolffd@0
|
84 % gpunpak - Separates hyperparameter vector into components.
|
|
wolffd@0
|
85 % gradchek - Checks a user-defined gradient function using finite differences.
|
|
wolffd@0
|
86 % graddesc - Gradient descent optimization.
|
|
wolffd@0
|
87 % gsamp - Sample from a Gaussian distribution.
|
|
wolffd@0
|
88 % gtm - Create a Generative Topographic Map.
|
|
wolffd@0
|
89 % gtmem - EM algorithm for Generative Topographic Mapping.
|
|
wolffd@0
|
90 % gtmfwd - Forward propagation through GTM.
|
|
wolffd@0
|
91 % gtminit - Initialise the weights and latent sample in a GTM.
|
|
wolffd@0
|
92 % gtmlmean - Mean responsibility for data in a GTM.
|
|
wolffd@0
|
93 % gtmlmode - Mode responsibility for data in a GTM.
|
|
wolffd@0
|
94 % gtmmag - Magnification factors for a GTM
|
|
wolffd@0
|
95 % gtmpost - Latent space responsibility for data in a GTM.
|
|
wolffd@0
|
96 % gtmprob - Probability for data under a GTM.
|
|
wolffd@0
|
97 % hbayes - Evaluate Hessian of Bayesian error function for network.
|
|
wolffd@0
|
98 % hesschek - Use central differences to confirm correct evaluation of Hessian matrix.
|
|
wolffd@0
|
99 % hintmat - Evaluates the coordinates of the patches for a Hinton diagram.
|
|
wolffd@0
|
100 % hinton - Plot Hinton diagram for a weight matrix.
|
|
wolffd@0
|
101 % histp - Histogram estimate of 1-dimensional probability distribution.
|
|
wolffd@0
|
102 % hmc - Hybrid Monte Carlo sampling.
|
|
wolffd@0
|
103 % kmeans - Trains a k means cluster model.
|
|
wolffd@0
|
104 % knn - Creates a K-nearest-neighbour classifier.
|
|
wolffd@0
|
105 % knnfwd - Forward propagation through a K-nearest-neighbour classifier.
|
|
wolffd@0
|
106 % linef - Calculate function value along a line.
|
|
wolffd@0
|
107 % linemin - One dimensional minimization.
|
|
wolffd@0
|
108 % maxitmess- Create a standard error message when training reaches max. iterations.
|
|
wolffd@0
|
109 % mdn - Creates a Mixture Density Network with specified architecture.
|
|
wolffd@0
|
110 % mdn2gmm - Converts an MDN mixture data structure to array of GMMs.
|
|
wolffd@0
|
111 % mdndist2 - Calculates squared distance between centres of Gaussian kernels and data
|
|
wolffd@0
|
112 % mdnerr - Evaluate error function for Mixture Density Network.
|
|
wolffd@0
|
113 % mdnfwd - Forward propagation through Mixture Density Network.
|
|
wolffd@0
|
114 % mdngrad - Evaluate gradient of error function for Mixture Density Network.
|
|
wolffd@0
|
115 % mdninit - Initialise the weights in a Mixture Density Network.
|
|
wolffd@0
|
116 % mdnpak - Combines weights and biases into one weights vector.
|
|
wolffd@0
|
117 % mdnpost - Computes the posterior probability for each MDN mixture component.
|
|
wolffd@0
|
118 % mdnprob - Computes the data probability likelihood for an MDN mixture structure.
|
|
wolffd@0
|
119 % mdnunpak - Separates weights vector into weight and bias matrices.
|
|
wolffd@0
|
120 % metrop - Markov Chain Monte Carlo sampling with Metropolis algorithm.
|
|
wolffd@0
|
121 % minbrack - Bracket a minimum of a function of one variable.
|
|
wolffd@0
|
122 % mlp - Create a 2-layer feedforward network.
|
|
wolffd@0
|
123 % mlpbkp - Backpropagate gradient of error function for 2-layer network.
|
|
wolffd@0
|
124 % mlpderiv - Evaluate derivatives of network outputs with respect to weights.
|
|
wolffd@0
|
125 % mlperr - Evaluate error function for 2-layer network.
|
|
wolffd@0
|
126 % mlpevfwd - Forward propagation with evidence for MLP
|
|
wolffd@0
|
127 % mlpfwd - Forward propagation through 2-layer network.
|
|
wolffd@0
|
128 % mlpgrad - Evaluate gradient of error function for 2-layer network.
|
|
wolffd@0
|
129 % mlphdotv - Evaluate the product of the data Hessian with a vector.
|
|
wolffd@0
|
130 % mlphess - Evaluate the Hessian matrix for a multi-layer perceptron network.
|
|
wolffd@0
|
131 % mlphint - Plot Hinton diagram for 2-layer feed-forward network.
|
|
wolffd@0
|
132 % mlpinit - Initialise the weights in a 2-layer feedforward network.
|
|
wolffd@0
|
133 % mlppak - Combines weights and biases into one weights vector.
|
|
wolffd@0
|
134 % mlpprior - Create Gaussian prior for mlp.
|
|
wolffd@0
|
135 % mlptrain - Utility to train an MLP network for demtrain
|
|
wolffd@0
|
136 % mlpunpak - Separates weights vector into weight and bias matrices.
|
|
wolffd@0
|
137 % netderiv - Evaluate derivatives of network outputs by weights generically.
|
|
wolffd@0
|
138 % neterr - Evaluate network error function for generic optimizers
|
|
wolffd@0
|
139 % netevfwd - Generic forward propagation with evidence for network
|
|
wolffd@0
|
140 % netgrad - Evaluate network error gradient for generic optimizers
|
|
wolffd@0
|
141 % nethess - Evaluate network Hessian
|
|
wolffd@0
|
142 % netinit - Initialise the weights in a network.
|
|
wolffd@0
|
143 % netopt - Optimize the weights in a network model.
|
|
wolffd@0
|
144 % netpak - Combines weights and biases into one weights vector.
|
|
wolffd@0
|
145 % netunpak - Separates weights vector into weight and bias matrices.
|
|
wolffd@0
|
146 % olgd - On-line gradient descent optimization.
|
|
wolffd@0
|
147 % pca - Principal Components Analysis
|
|
wolffd@0
|
148 % plotmat - Display a matrix.
|
|
wolffd@0
|
149 % ppca - Probabilistic Principal Components Analysis
|
|
wolffd@0
|
150 % quasinew - Quasi-Newton optimization.
|
|
wolffd@0
|
151 % rbf - Creates an RBF network with specified architecture
|
|
wolffd@0
|
152 % rbfbkp - Backpropagate gradient of error function for RBF network.
|
|
wolffd@0
|
153 % rbfderiv - Evaluate derivatives of RBF network outputs with respect to weights.
|
|
wolffd@0
|
154 % rbferr - Evaluate error function for RBF network.
|
|
wolffd@0
|
155 % rbfevfwd - Forward propagation with evidence for RBF
|
|
wolffd@0
|
156 % rbffwd - Forward propagation through RBF network with linear outputs.
|
|
wolffd@0
|
157 % rbfgrad - Evaluate gradient of error function for RBF network.
|
|
wolffd@0
|
158 % rbfhess - Evaluate the Hessian matrix for RBF network.
|
|
wolffd@0
|
159 % rbfjacob - Evaluate derivatives of RBF network outputs with respect to inputs.
|
|
wolffd@0
|
160 % rbfpak - Combines all the parameters in an RBF network into one weights vector.
|
|
wolffd@0
|
161 % rbfprior - Create Gaussian prior and output layer mask for RBF.
|
|
wolffd@0
|
162 % rbfsetbf - Set basis functions of RBF from data.
|
|
wolffd@0
|
163 % rbfsetfw - Set basis function widths of RBF.
|
|
wolffd@0
|
164 % rbftrain - Two stage training of RBF network.
|
|
wolffd@0
|
165 % rbfunpak - Separates a vector of RBF weights into its components.
|
|
wolffd@0
|
166 % rosegrad - Calculate gradient of Rosenbrock's function.
|
|
wolffd@0
|
167 % rosen - Calculate Rosenbrock's function.
|
|
wolffd@0
|
168 % scg - Scaled conjugate gradient optimization.
|
|
wolffd@0
|
169 % som - Creates a Self-Organising Map.
|
|
wolffd@0
|
170 % somfwd - Forward propagation through a Self-Organising Map.
|
|
wolffd@0
|
171 % sompak - Combines node weights into one weights matrix.
|
|
wolffd@0
|
172 % somtrain - Kohonen training algorithm for SOM.
|
|
wolffd@0
|
173 % somunpak - Replaces node weights in SOM.
|
|
wolffd@0
|
174 %
|
|
wolffd@0
|
175 % Copyright (c) Ian T Nabney (1996-2001)
|
|
wolffd@0
|
176 %
|
