MODULAR & READY-TO-USE MACHINE LEARNING ENGINE
SignalReveal is a fully modular Machine Learning engine to create signals and investment strategies from any traditional market, alternative, or proprietary dataset.
Its models and Python source code are fully transferred to clients.
This powerful tool accelerates data analysis and provides techniques to protect against overfitting, all with a transparent, no black box approach.
Alternative Data Sets Evaluation & Integration
Rapidly and rigorously evaluate any structured dataset and integrate it with other datasets
Modular Machine Learning Pipelines
Build predictive machine learning pipelines to identify sources of alpha and limit the risk of overfitting
Signal & Trading Strategy Creation
Build predictive indicators and complete trading strategies for any asset class or use case
A transparent & standardized Machine Learning methodology to help you gain time.
- Alternative Data
- Market Data
- Quantitative factors
- Explanatory & target features
- Correlations & Stationarity
- Discriminating importance
- Dimension Reduction
- Machine Learning & Quantitative Models
- Temporal Cross-Validation
- Models combination & interpretation
- Stability checks
WORKING WITH SESAMm
Get access to technical documentation & Notebooks to evaluate SignalReveal.
Dive deeper with our team and ask your technical questions during a Discovery Workshop.
3. Code Delivery
Receive SignalReveal’s Python source code, documentation & Notebooks for each module.
4. Training & Support
Training, support and versioning are included. Be operational in just a few days.
Model Selection Notebook Sample
- Dimension reduction;
- Graphical model selection.
Notebook sample: model selection with SignalReveal for a long-only and long-short equity use case combining alternative and market data.
Please use Chrome or Firefox for optimal display of this Notebook.