Ads by Google
Christian Borgelt's Web Pages

Accretion - Finding Neuronal Assemblies

Download

32 bit 64 bit (32/64 bit only for executable)
accretion (258 kb) accretion (280 kb) GNU/Linux executable
accretion.exe (182 kb) accretion.exe (213 kb) Windows console executable
accretion.zip (165 kb) accretion.tar.gz (148 kb) C sources, version 2.18 (2017.05.30)

Description

Accretion is a program to execute the Accretion algorithm ([Gerstein et al. 1978], see below) on neural spike train data. Since the Accretion algorithm is closely related to frequent item set mining, the program uses the terminology of this area. That is, the program refers to items rather than neurons and to transactions rather than time bins.

Scripts to convert the input from GDF format (two column format consisting of time stamp and neuron id) to the needed transaction format can be found in the directory accretion/ex in the source package.

The implementation uses an Eclat-based scheme to determine the needed support counters, but prunes the search mainly with the results of statistical tests. (Compare the additional evaluation measures for the Apriori program.)

Note that due to the processing scheme of the Accretion algorithm the output may contain the same set of items/neurons multiple times with the same or different statistical tests results. This is not a bug of the program, but a drawback of the Accretion algorithm.

This implementation may also be used through the Python interface provided by the PyFIM library.

If you have trouble executing the program on Microsoft Windows, check whether you have the Microsoft Visual C++ Redistributable for Visual Studio 2017 (see under "Other Tools and Frameworks") installed, as the program was compiled with Microsoft Visual Studio 2017.

A comparison of the Accretion algorithm to frequent item set mining can be found in:

Reference for the original Accretion algorithm: