Normalizing Price Angles for Machine Learning Market Analysis

Summary

This article proposes treating consecutive price changes as angles and using sequences of those angles as features for market prediction. It frames the idea as a more systematic alternative to manually drawing Gann angles, with the goal of capturing direction and slope changes. The described workflow retrieves EUR/USD data from MetaTrader 5, scales the time and price axes to comparable ranges, calculates angles between successive observations, and applies machine learning to identify patterns associated with later price moves.

The author reports observing sequences of negative, near-neutral, and increasingly positive angles before some upward moves, and says a pattern near a 45-degree rise echoed a classical Gann idea. The discussion refers to months of observation on EUR/USD and presents qualitative claims rather than detailed validation statistics. It does not specify enough about the predictive model, training and test separation, or out-of-sample performance to assess whether the patterns generalize. Angle values also depend on the normalization choices, so the method requires careful scaling and independent testing before use as a trading signal.

Key ideas

• The method measures the angle between successive prices after scaling time and price to comparable ranges.
• Sequences of angle features are proposed as inputs for machine learning models.
• The article relates the approach to Gann analysis while replacing manually drawn lines with calculated observations.
• Reported patterns come from qualitative EUR/USD observations, including examples before upward moves.
• The document does not provide sufficient model or out-of-sample details to establish predictive reliability.

Tags

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