Playground for machine learning with medical coding data and medical classifications.
Tasks
- propose codes (denoising, auto encoding, recommender systems, Analog collaborative filtering / matrix factorization, clasisification, sequence mapping)
- learn a grouper algorithm (classification, outlier detection, sequence mapping)
- search codes (medical classifications, NLP)
- exploration and visualisation (Visualised Bayes Nets, common combinations of comorbidities)
- medical coding data generation
- free text analysis of medical documentations (generate code proposals)
- ..
Algorithms
- classical machine learning algorithms (FFNNs, RF, PCA, RBM)
- deep learning (especially LSTMs)
- word2vec
- ..
Libraries
- scikit-learn
- Keras
- Theano
- TensorFlow
- nltk
- ..
Links
- Word2Vec explained: http://www-personal.umich.edu/~ronxin/pdf/w2vexp.pdf