/TF_library

TensorFlow Models

Primary LanguageR

TensorFlow Model and ML Project Library

A collection of ML models built with the TensorFlow api package for python3 (https://www.tensorflow.org) and other modeling tools, such as glmnet package in R.

Models are in the main directory, while training and testing data is in the /data repo.

Model Types:

  • #f03c15 TensorFlow
  • #1589F0 glmnet

Model Collection:

  1. #f03c15 TensorFlow: Diabetes Patient Classification: application of the softmax regression model to group patients not diagnosed with diabetes, from those with either Chemical or Overt diabetes, based on several clinical measures. Training and testing obtained from the R dataset 'Diabetes' from the 'heplots' library (Original data from Reaven and Miller (1979)).

  2. #1589F0 glmnet: LIHC Tumor Stage Classification using Viral Load: application of multinomial regression in glmnet to create a model predicting patient tumor stage classification based on HBV and total viral load, as reported in by Cao et. Al (Cao, S et. al. 'Divergent viral presentation among human tumors and adjacent normal tissues'. Sci. Rep., 6,28294 (2016)). Tumor stage and clinical information retrieved from TCGA biobank (All LIHC patients).