Skip to content
This repository has been archived by the owner on Jan 9, 2025. It is now read-only.

Method to use rolling optimization with 3D LSTM arrays. Explores integrating multi-asset portfolio with embedding layer.

License

Notifications You must be signed in to change notification settings

DeepNexus1/walk-forward-lstm-3d

Repository files navigation

Deep Nexus Research Repository for Categorical LSTM with Walk-Forward Optimization

Project Overview

This repository contains two LSTM Sequence-to-Sequence model examples with rolling walk-forward optimization. The walk-forward optimization is commonly used with financial time-series data to address market regime-change.

Each model uses categorical outputs and can be highly customized to provide signals for binary, ternary, or other ranking outputs.

One of the models is adapted for a portfolio of assets and utilizes an Embedding Layer to learn relationships between assets.

Core Components

  • Prediction Models:
    • Implemented using TensorFlow and Keras
    • Long Short-Term Memory (LSTM) neural network architecture
    • Categorical Prediction Output

Technical Objectives

  • Ensuring 3-D arrays are properly structured for walk-forward training
  • Flexible structure for a variety of categorical outputs
  • Capacity for feature creation and scaling at sequence generation
  • Out-of-Sample Performance Evaluation Metrics
  • Accommodating multiple assets in one predictive model
  • Implement Embedding Layer in Portfolio model

Walk-Forward and Tensor Illustration

Figure: Walk-Forward Training (left) and Portfolio Batch Illustration (right).

Limitations and Considerations

  • Research project from 2020.
  • Python 3.5 was used to build this project.
  • Uses basic input features and output categories intended for testing purposes only.
  • Must be customized with suitable features and prediction targets.
  • Import your own data to train the model, as data is not provided.
  • Final production models are proprietary.

Disclaimer

This is a research project and should not be considered financial advice. Trading involves significant financial risk.

License

http://www.apache.org/licenses/LICENSE-2.0

Contact

web@deepnexus.com

Repository initialized in 2025, based on research conducted in 2020.

About

Method to use rolling optimization with 3D LSTM arrays. Explores integrating multi-asset portfolio with embedding layer.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages