mudafar / brain-cemisid-parallel

BrainCEMISID V2.1

The BrainCEMISID project is the University of Los Andes (Mérida-Venezuela) first cognitive architecture that tackles problems such as visual and auditive information representation, sensorial ambiguities solving, basic visual stimula composition into more complex cultural information, autonomous intentions-based decision making, among others.

New in this version:

TO-DO:

• Make all of the neuronal networks parallel

Installation

The BrainCEMISID kernel is written in Python 2.7. This means that all machines with a 2.7 Python interpreter can run the code.

###Kernel dependencies:

• Drill: download and follow installation instructions https://github.com/uqfoundation/dill

• Python multiprocessing library.

UI (User interface) demo dependencies:

• Kivy 1.9.1, available at https://kivy.org/#download. Follow the installation instructions.

###Running: python main.py

Most of the modules, such as the MulticlassSingleLayerNetwork, contain independent tests that can be of help for the comprehension of their functionality. Run, for instance: python multiclass_single_layer_network.py

Contributors

• Jonathan Monsalve
• Julio Muchacho
• Ricardo Graterol
• Ricardo Bruzual
• Johan Sosa
• Breytner Fernández, ([email protected])
• Mudafar El Halabi