Entropy-Assisted Identification of Predictive Short-Term Price Patterns

Summary

This paper proposes a framework for finding short-term patterns in financial time series that may signal future price direction. It seeks high-quality, non-overlapping patterns by clustering historical examples and pruning less useful ones. The method uses local entropy as a proxy for information content: patterns associated with strongly one-sided historical movements and low local entropy are treated as more informative. It also aims to retain a balanced representation of Buy and Sell patterns.

The authors contrast this approach with K-means and Gaussian Mixture Models, arguing that conventional clustering can create biased or unbalanced groups and lose useful patterns through over-segmentation. Their framework gives weight to both predictive purity and historical profitability. The document describes the rationale and intended advantage, but supplies no dataset, quantitative results, validation design, or live trading evidence. Its claims about predictive power and trading suitability therefore remain unverified by the information provided.

Key ideas

• The framework searches for non-overlapping short-term patterns in noisy financial time series.
• Local entropy is used as a proxy for the information content of a historical pattern.
• Patterns linked to one-sided historical movement and low local entropy are prioritized.
• The approach seeks balanced Buy and Sell groups and compares itself with K-means and Gaussian mixture clustering.
• The document provides no quantitative validation or evidence of live trading results.

Tags

# Entropy-Assisted Quality Pattern Identification in Finance


# Entropy-Assisted Quality Pattern Identification in Finance









Short-term patterns in financial time series form the cornerstone of many algorithmic trading strategies, yet extracting these patterns reliably from noisy market data remains a formidable challenge. In this paper, we propose an entropy-assisted framework for identifying high-quality, non-overlapping patterns that exhibit consistent behavior over time. We ground our approach in the premise that historical patterns, when accurately clustered and pruned, can yield substantial predictive power for short-term price movements. To achieve this, we incorporate an entropy-based measure as a proxy for information gain. Patterns that lead to high one-sided movements in historical data, yet retain low local entropy, are more informative in signaling future market direction. Compared to conventional clustering techniques such as K-means and Gaussian Mixture Models (GMM), which often yield biased or unbalanced groupings, our approach emphasizes balance over a forced visual boundary, ensuring that quality patterns are not lost due to over-segmentation. By emphasizing both predictive purity (low local entropy) and historical profitability, our method achieves a balanced representation of Buy and Sell patterns, making it better suited for short-term algorithmic trading strategies.

Shown in full with attribution under the source's licence. Licence: abstract CC0

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