Neural-Nets-Final-Project

Project by Kunal Patel, Nathan Pilbrough, Priscilla Cheng

UCLA EE 239AS: Neural Networks (by Prof. Jonathon Kao)

Python 2.7.12

Dataset

-Witheld for confidentiality purposes and owner rights

Library Dependencies

Execute "pip install -r requirements.tct" within your virtual environment

Code Execution

All models are implemented within the following files
-RecurrentModels.ipynb
-BraindecodeTest.ipynb
-DeepConvNet.ipynb
-DeepConvNetAll.ipynb
-DeepConvNetSTFT.ipynb
Implemented functions are found in UtilNNDL.py

Save Data Naming Standard (./Best Models/Variables)

hist.History dictionaries:
[modelname]_hist_[SubjectName].npy

test_accuracy variable:
[modelname]_testacc_[SubjectName].npy

Confusion_matrix variable:
[modelname]_conf_[SubjectName].npy

Avoid underscores in modelname, makes the information easier to tokenize if we need to in the future
SubjectName examples = 'All', 'A10T', etc.

Resources:

https://kupdf.com/download/long-short-term-memory-networks-with-python_5a43310ce2b6f5d926656665_pdf
http://www.bbci.de/competition/iv/desc_2a.pdf
https://arxiv.org/pdf/1802.00308.pdf - ChronoNet