A Simple Backpropagation Network for Charting Short-Term Price Projections

Summary

The document describes a compact artificial neural network with one hidden layer, trained using backpropagation, and a chart indicator that visualizes price movement while projecting future closes. The indicator uses recent open prices as inputs, compares its output with the last open price, and adjusts network weights whenever a candle completes. Its default configuration uses 200 past opens and projects 10 closes ahead.

This is presented as an educational demonstration of a neural network adapting as new market data arrives, not as a tested forecasting strategy. The author explicitly cautions that the output should be treated as an experimental opinion rather than a dependable prediction or trading signal. The document gives no accuracy measures, benchmark comparisons, or trading results, so it does not establish that the projected values have predictive or economic value. Both described files are said to work without external library dependencies.

Key ideas

• The example uses a single-hidden-layer neural network trained with backpropagation.
• It updates network weights as each completed candle arrives.
• Recent open prices are used to project a configurable sequence of future closes.
• The author presents the output as an educational visualization rather than a reliable forecast.
• No accuracy statistics, benchmarks, or trading results are supplied.

Tags

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