Download Computation and Neural Systems by Anthony J. Bell (auth.), Frank H. Eeckman, James M. Bower PDF

By Anthony J. Bell (auth.), Frank H. Eeckman, James M. Bower (eds.)

Computational neuroscience is healthier outlined via its specialize in knowing the apprehensive platforms as a computational equipment instead of by way of a specific experimental approach. Accordinlgy, whereas nearly all of the papers during this ebook describe research and modeling efforts, different papers describe the result of new organic experiments explicitly put within the context of computational concerns. The distribution of matters in Computation and Neural Systems displays the present country of the sphere. as well as the clinical effects awarded right here, a number of papers additionally describe the continued technical advancements which are severe for the continuing progress of computational neuroscience.
Computation and Neural Systems comprises papers provided on the First Annual Computation and Neural structures assembly held in San Francisco, CA, July 26--29, 1992.

Show description

Read or Download Computation and Neural Systems PDF

Similar nonfiction_8 books

Document Processing and Retrieval: Texpros

Record Processing and Retrieval: TEXPROS makes a speciality of the layout and implementation of a private, customizable place of work info and record processing approach referred to as TEXPROS (a textual content PROcessing System). TEXPROS is a private, clever workplace info and rfile processing method for text-oriented records.

Anisotropy Effects in Superconductors

Contemplating the big growth accomplished in a few distinctive components of superconductivity over the past few years, it appeared beneficial to debate completely a topic, which has encountered serious seasoned­ blems at the theoretical and the experimental aspect, specifically the consequences of the anisotropic electron and phonon homes of (single crystalline) fabrics at the attribute gains of the great­ carrying out country.

Additional resources for Computation and Neural Systems

Example text

2 The PSP of the neuron after learning when a pattern in input versus the correlation between patterns. Here b = 2, there are two patterns, and c• = 1/3. The dashed curve shows the unstable fixed point. 3 Simulation of learning. (a) Six patterns presented to the neuron, consisting of 4096 pixels. (b) The synaptic weight vector after 0, 80, 84, 86, 90 and 100 learning steps forb= 2. The neuron has learned to recognize U in the presence of other patterns. 3. The neuron has learned a mixture of the letters.

E. the principal component) of the correlation matrix Mij =< XiXj >, where < ... > denotes average over the patterns. This has an important statistical interpretation - it is the first stage of principal component analysis (PCA), which is a well known method of data compression. Using these neurons as building blocks, Oja and other researchers have developed neural network architectures which perform principal component analysis [1, 4]. Because the linear neuron learns the principal component of the correlation matrix which is a statistical property of the ensemble of patterns, it learns about the pattern set, not about individual patterns.

I and y are the unit vectors. The angle of this vector is given by the arctangent of the ratio of the postcross interval and the pre-cross interval. The scalar quantity of this cross-interval vector, v, is analogous to the first-order "conditional cross-interval" measure introduced by Tam et al [5]. Given n spike trains, there are a total of n- 1 such cross-interval vectors for any given pair of neurons. A resultant vectorial sum, V, of these n- 1 vectors can be obtained for each given reference spike: n-1 n-1 i=1 i=l v =L, v; == L,ul+ J;,) <8-2) where v; denotes the cross-interval vector with respect to the i-th spike train.

Download PDF sample

Rated 4.30 of 5 – based on 35 votes