Combining ADX Trend Strength and CCI Momentum Patterns with an MLP

Summary

This article explains how ADX and CCI can be paired as inputs to a supervised multilayer perceptron. ADX measures trend strength without indicating direction, while CCI tracks price momentum relative to a typical-price average. It outlines Python calculations for both indicators, including Wilder-style smoothing for ADX, and discusses data requirements and possible issues such as missing initial values, zero denominators, and the cost of calculating CCI mean deviation.

The described signals combine ADX strength with CCI moves around zero and prior threshold levels. The models were evaluated in a forward walk, but the tested patterns did not produce results suitable for sharing, and the article characterizes the overall findings as mixed to poor. It offers no evidence of a profitable strategy. The indicator code and signal definitions are implementation examples; the article notes that its continuous input approach may have contributed to weak results and proposes exploring other machine-learning methods in later work.

Key ideas

• ADX estimates trend strength, while CCI represents momentum relative to a typical-price average.
• The article computes CCI from high, low, and close prices, then scales deviations from a moving average by mean deviation.
• Its ADX implementation derives directional movement and true range, smooths them, and calculates directional indicators and ADX.
• The proposed signal patterns combine ADX strength with CCI moves around zero and earlier threshold readings.
• The reported forward-walk results were mixed to poor, so the patterns are not established as effective trading signals.

Tags

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