Finding Similar Candlestick Patterns with Fuzzy Matching

Summary

This document outlines a research tool for recording custom single-candle patterns and searching historical or current data for similar examples. A pattern is represented by open, high, low, and close values, with an optional label such as a bullish engulfing pattern. The user can set a tolerance so the detector accepts approximate matches instead of requiring every price value to be identical.

The idea is useful for exploratory pattern research, but the document does not specify how values are normalized, how the fuzzy distance is calculated, or how matches are ranked. It gives no evidence that similar candle shapes have predictive value, and an example tolerance is illustrative rather than validated. Researchers would need to define matching rules carefully and test candidate patterns out of sample, while accounting for the possibility that flexible matching finds coincidental similarities.

Key ideas

• A custom candle pattern can be stored using its open, high, low, and close values and a label.
• Fuzzy matching allows later candles to differ from the stored example within a chosen tolerance.
• The document does not define the matching metric or normalization method.
• Finding similar shapes alone does not establish predictive value; patterns require empirical testing.

Tags

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