Skip to content

Latest commit

 

History

History
3 lines (2 loc) · 488 Bytes

README.md

File metadata and controls

3 lines (2 loc) · 488 Bytes

LSTM

All purpose OOP implementation of a Long Short-Term Memory Recurrent Neural Network, using TensorFlow. The model was initially produced for my Undergraduate Thesis in predicting stock price movements. It has been designed in a highly encapsulated form (including completely hiding the TensorFlow Session within the class) with many user-customisable parameters defining the network's architecture, allowing it to be easily adjusted and applied to any time-series prediction task.