TensorFlow-based implementations of several Machine Learning models (first three - Logistic Regresion, MLP, and CNN - are heavily inspired by TensorFlow v1.3 tutorials). The models folder contains simple implementations of:
- Logistic Regression
- Multi-Layer Perceptron
- Convolutional Neural Network
- K-Means Clustering
- Gaussian Mixture Model (with EM)
The gmm folder contains more elaborate versions of a Gaussian Mixture Model implementation trained by means of Expectation Maximization algorithm (with diagonal covariance, full covariance, gradient-based, etc.). The gmm/struct folder contains initial attempts to decompose the GMM implementation into a coherent set of classes.