Skip to content
All library documents

DoubleEnsemble: Reweight Samples and Select Features for Market Prediction

Article Qlib

Summary

DoubleEnsemble is an ensemble framework for financial prediction that addresses noisy training data and large feature sets. It combines two procedures: it uses each sample’s learning trajectory during training to identify influential examples and adjust their weights, and it evaluates features by shuffling them to measure their impact on model performance. The selected samples and features are then used with a base model to capture patterns while seeking to reduce overfitting and instability.

The description presents the approach as compatible with a range of base models and points to a research paper on financial data analysis. It does not report specific datasets, validation results, benchmark comparisons, or implementation details here, so its claimed benefits cannot be assessed from this summary alone. In practice, the method’s usefulness depends on how sample weighting and feature selection are performed and validated against leakage and changing market conditions.

Key ideas

  • DoubleEnsemble combines sample reweighting with feature selection for financial prediction.
  • It uses training dynamics to identify influential samples.
  • It assesses feature importance by shuffling features and measuring the resulting impact.
  • The framework is designed to work with different base models, but the document gives no benchmark results.

Tags

Machine learningStatistics
Full text
# DoubleEnsemble


# DoubleEnsemble
* DoubleEnsemble is an ensemble framework leveraging learning trajectory based sample reweighting and shuffling based feature selection, to solve both the low signal-to-noise ratio and increasing number of features problems. They identify the key samples based on the training dynamics on each sample and elicit key features based on the ablation impact of each feature via shuffling. The model is applicable to a wide range of base models, capable of extracting complex patterns, while mitigating the overfitting and instability issues for financial market prediction.
* This code used in Qlib is implemented by ourselves.
* Paper: DoubleEnsemble: A New Ensemble Method Based on Sample Reweighting and Feature Selection for Financial Data Analysis [https://arxiv.org/pdf/2010.01265.pdf](https://arxiv.org/pdf/2010.01265.pdf).

Shown in full with attribution under the source's licence. Licence: MIT

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.