Multi-Timeframe Moving Average Signals with a Tunable Neural Unit

Summary

The document describes an Expert Advisor that combines moving-average changes from hourly, four-hour, and daily charts. Each timeframe contributes three relative changes between recent moving-average values, weighted by adjustable parameters and rounded after scaling. The resulting values act as simple signal units: aligned positive readings across all three timeframes trigger a buy condition, while a positive hourly reading alongside negative four-hour and daily readings triggers a sell condition.

The author recommends optimizing parameters in two stages, then choosing from the top candidates those that perform similarly in backtests and forward tests. No performance figures, detailed optimization procedure, or market and instrument specifics are provided. Although the unit is called a neural network, the excerpt shows a weighted calculation and threshold rules rather than a described training process or learned network architecture. The signals also depend on moving-average changes and parameter selection, so the document does not establish whether they are profitable or robust across markets.

Key ideas

• The signal uses weighted relative changes in moving averages from hourly, four-hour, and daily timeframes.
• Positive readings across all three timeframes indicate a buy condition in the described rules.
• A positive hourly reading with negative four-hour and daily readings indicates a sell condition.
• The author suggests staged parameter optimization and comparing backtest with forward-test results.
• The document provides no performance evidence or details of a neural-network training process.

Tags

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