deep-learning

Implementations of Deep Neural Networks, Recurrent Neural Networks and Convolutional Neural Networks and other deep learning architectures and techniques

Description of files:

• tf_acoustic_embedding: Acoustic embeddings on the TIMIT dataset using Variational Autoencoders
• tf_cnn: Convolution Neural Network teseted on notMNIST dataset
• tf_sentiment_analyzer: Sentiment analysis using Convolution Neural Networks
• tf_deep_neural_network: Deep Neural Network tested on (Oh God! Not this dataset again) MNIST dataset
• tf_gan: Generative Adversaria Network to generate (wait for it) MNIST digits
• tf_recurrent_neural_network: Recurrent Neural Network to generate text
• theano_dnn: Deep Neural Network implementation in Theano
• theano_rnn: Vanilla Neural Network implementation in Theano
• theano_sparse_autoencoder: Sparse autoencoder implementation in Theano
• gists: Gists of video lectures/research papers
• notebooks: iPython notebooks for deep learning

PS: Most of the code in here were implemented quite sometime ago and have not been updated. Thus, a lot of backward compatability might be broken (especially with TensorFlow..)