Developing a trading strategy for Bitcoin Market using Long Short-term Memory (LSTM) architecture

Prediction of financial markets time series market direction is a challenging task mainly due to the unprecedented changes in economic trends and conditions in one hand and incomplete information on the other hand. Therefore, developing different forecasting, like LSTMs, have been employed by quantitative traders recently.
Long Short Term Memory networks (LSTMs”) – are a special kind of recurrent neural network (RNN), capable of learning long-term dependencies. LSTMs are explicitly designed to avoid the long-term dependency problem, with additional features to memorize the sequence of data.
The purpose of this project is to use LSTMs to develop trading and day-trading strategies for Bitcoin Spot market. The inputs to the LSTM models are found through a supervised feature selection process over a pool of various features. These features include a set of technical indicators, and another feature obtained from market sentiment analysis.

Faculty Supervisor:

Michèle Breton

Student:

Partner:

Pow.re Corporation

Discipline:

Mathematics

Sector:

Finance and Insurance

University:

HEC Montréal

Program: