Combining FrAMA and Force Index Patterns with a Dot Product CNN
Summary
The article describes a machine learning extension to an earlier Expert Advisor using Fractal Adaptive Moving Average (FrAMA) and Force Index signals. It outlines Python implementations of both indicators: FrAMA adapts EMA smoothing using a simplified estimate of price fractal dimension, while Force Index weights price changes by volume and smooths them with an EMA. It then applies a convolutional neural network using a dot product kernel to selected signal patterns.
Of ten previously optimized patterns, the article focuses on patterns 6 and 9 because those were the only ones reported to forward walk. Both are said to continue forward walking in the neural network runs, but the text provides no detailed performance figures in the excerpt. The author emphasizes that training used a limited dataset and warns that broader historical testing is needed before drawing conclusions. The simplified indicator formulas also require care around initialization, missing values, and data alignment.
Key ideas
- FrAMA adjusts its EMA smoothing according to a simplified fractal dimension estimate of price movement.
- Force Index combines price changes and volume, then smooths the result with an EMA.
- A dot product kernel CNN is applied to two previously selected indicator signal patterns.
- The reported forward walks are preliminary because training used limited data and requires broader testing.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.