Skip to content

A simple deep learning approach is present to predict the direction of high-frequency stock price changes.

Notifications You must be signed in to change notification settings

niklaswalter/Binary_Classifier_Stock_Movements

Repository files navigation

Binary_Classifier_Stock_Movements

This notebook presents a simple deep learning approach to predict movements of a stock. It is based on high-frequency limit order book data and information about former movements. Based on this and the order imbalance we build a fully-connected neural network to run a prediction on an unlabelled dataset.

For the implementation of the model we use the Keras library.

About

A simple deep learning approach is present to predict the direction of high-frequency stock price changes.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published