Skip to content

Performance Analysis of Trading Rules Via Artificial Neural Networks in Trading Operations.

License

Notifications You must be signed in to change notification settings

viniciusbarbosapaiva/Mestrado

Repository files navigation

Project Title

Performance Analysis of Trading Rules Via Artificial Neural Networks in Trading Operations.

Description

This project proposes to measure the accuracy of the trading rule known as moving average crossover. For this task, the application of the Deep Artificial Neural Network (DNN) model based on the binary classification was proposed. The financial asset obtained for the research was the Mini U.S. Dollar Futures Contract, as of the second half of 2019. For the DNN model, 22 variables were defined. Among these: the closing price, opening price, maximum price and minimum price. Collected from a classification robot, such variables are input parameters for the proposed DNN architecture.

Getting Started

Dependencies

  • numpy
  • pandas
  • matplotlib
  • warnings
  • calendar
  • seaborn
  • sklearn
  • tensorflow
  • keras

Project Results

The results discussed in this project were forwarded at the 40th International Forecasting Symposium (ISF) and Production Engineering Symposium (SIMEP). With the data and models used, the final results defined that the accuracy found for the trading rule of crossing moving averages, from the DNN model, presented a low percentage of prediction.

About

Performance Analysis of Trading Rules Via Artificial Neural Networks in Trading Operations.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published