Results 1 
7 of
7
Fast Exact Multiplication by the Hessian
 Neural Computation
, 1994
"... Just storing the Hessian H (the matrix of second derivatives d^2 E/dw_i dw_j of the error E with respect to each pair of weights) of a large neural network is difficult. Since a common use of a large matrix like H is to compute its product with various vectors, we derive a technique that directly ca ..."
Abstract

Cited by 71 (4 self)
 Add to MetaCart
Just storing the Hessian H (the matrix of second derivatives d^2 E/dw_i dw_j of the error E with respect to each pair of weights) of a large neural network is difficult. Since a common use of a large matrix like H is to compute its product with various vectors, we derive a technique that directly calculates Hv, where v is an arbitrary vector. This allows H to be treated as a generalized sparse matrix. To calculate Hv, we first define a differential operator R{f(w)} = (d/dr)f(w + rv)_{r=0}, note that R{grad_w} = Hv and R{w} = v, and then apply R{} to the equations used to compute grad_w. The result is an exact and numerically stable procedure for computing Hv, which takes about as much computation, and is about as local, as a gradient evaluation. We then apply the technique to backpropagation networks, recurrent backpropagation, and stochastic Boltzmann Machines. Finally, we show that this technique can be used at the heart of many iterative techniques for computing various properties of H, obviating the need for direct methods.
Computing Second Derivatives in FeedForward Networks: a Review
 IEEE Transactions on Neural Networks
, 1994
"... . The calculation of second derivatives is required by recent training and analyses techniques of connectionist networks, such as the elimination of superfluous weights, and the estimation of confidence intervals both for weights and network outputs. We here review and develop exact and approximate ..."
Abstract

Cited by 28 (4 self)
 Add to MetaCart
. The calculation of second derivatives is required by recent training and analyses techniques of connectionist networks, such as the elimination of superfluous weights, and the estimation of confidence intervals both for weights and network outputs. We here review and develop exact and approximate algorithms for calculating second derivatives. For networks with jwj weights, simply writing the full matrix of second derivatives requires O(jwj 2 ) operations. For networks of radial basis units or sigmoid units, exact calculation of the necessary intermediate terms requires of the order of 2h + 2 backward/forwardpropagation passes where h is the number of hidden units in the network. We also review and compare three approximations (ignoring some components of the second derivative, numerical differentiation, and scoring). Our algorithms apply to arbitrary activation functions, networks, and error functions (for instance, with connections that skip layers, or radial basis functions, or ...
Bayesian Regularisation and Pruning using a Laplace Prior
 Neural Computation
, 1994
"... Standard techniques for improved generalisation from neural networks include weight decay and pruning. Weight decay has a Bayesian interpretation with the decay function corresponding to a prior over weights. The method of transformation groups and maximum entropy indicates a Laplace rather than a G ..."
Abstract

Cited by 18 (0 self)
 Add to MetaCart
Standard techniques for improved generalisation from neural networks include weight decay and pruning. Weight decay has a Bayesian interpretation with the decay function corresponding to a prior over weights. The method of transformation groups and maximum entropy indicates a Laplace rather than a Gaussian prior. After training, the weights then arrange themselves into two classes: (1) those with a common sensitivity to the data error (2) those failing to achieve this sensitivity and which therefore vanish. Since the critical value is determined adaptively during training, pruningin the sense of setting weights to exact zerosbecomes a consequence of regularisation alone. The count of free parameters is also reduced automatically as weights are pruned. A comparison is made with results of MacKay using the evidence framework and a Gaussian regulariser. 1 Introduction Neural networks designed for regression or classification need to be trained using some form of stabilisation or re...
Efficient Training of FeedForward Neural Networks
, 1997
"... : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 61 A.2 Introduction : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 61 A.2.1 Motivation : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 61 A.3 Optimization strategy : : : : : : : : : : : : ..."
Abstract

Cited by 12 (0 self)
 Add to MetaCart
: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 61 A.2 Introduction : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 61 A.2.1 Motivation : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 61 A.3 Optimization strategy : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 62 A.4 The Backpropagation algorithm : : : : : : : : : : : : : : : : : : : : : : : : 63 A.5 Conjugate direction methods : : : : : : : : : : : : : : : : : : : : : : : : : : 63 A.5.1 Conjugate gradients : : : : : : : : : : : : : : : : : : : : : : : : : : 65 A.5.2 The CGL algorithm : : : : : : : : : : : : : : : : : : : : : : : : : : : 67 A.5.3 The BFGS algorithm : : : : : : : : : : : : : : : : : : : : : : : : : : 67 A.6 The SCG algorithm : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 67 A.7 Test results : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 70 A.7.1 Comparison metric : : : : : : : : : : : : : : : : : : : : : : : :...
Flat Minimum Search Finds Simple Nets
, 1994
"... We present a new algorithm for finding low complexity neural networks with high generalization capability. The algorithm searches for a "flat" minimum of the error function. A flat minimum is a large connected region in weightspace where the error remains approximately constant. An MDLbased argume ..."
Abstract

Cited by 3 (2 self)
 Add to MetaCart
We present a new algorithm for finding low complexity neural networks with high generalization capability. The algorithm searches for a "flat" minimum of the error function. A flat minimum is a large connected region in weightspace where the error remains approximately constant. An MDLbased argument shows that flat minima correspond to low expected overfitting. Although our algorithm requires the computation of second order derivatives, it has backprop's order of complexity. Automatically, it effectively prunes units, weights, and input lines. Various experiments with feedforward and recurrent nets are described. In an application to stock market prediction, flat minimum search outperforms (1) conventional backprop, (2) weight decay, (3) "optimal brain surgeon" / "optimal brain damage".
Simplifying Neural Nets by Discovering Flat Minima
 Advances in Neural Information Processing Systems 7
, 1995
"... We present a new algorithm for finding low complexity networks with high generalization capability. The algorithm searches for large connected regions of socalled "fiat" minima of the error function. In the weightspace environment of a "fiat" minimum, the error remains approximately constant. U ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
We present a new algorithm for finding low complexity networks with high generalization capability. The algorithm searches for large connected regions of socalled "fiat" minima of the error function. In the weightspace environment of a "fiat" minimum, the error remains approximately constant. Using an MDLbased argument, fiat minima can be shown to correspond to low expected overfitting. Although our algorithm requires the computation of second order derivatives, it has backprop's order of complexity. Experiments with feedforward and recurrent nets are described.