Dual-Timeframe Neural Network Signals for Trend Trading

Summary

This strategy combines predictions from a neural network applied to price changes on a large and a small timeframe. It opens a long or short position only when both model outputs exceed a threshold in the same direction, and closes positions when the predictions no longer agree. The example settings use daily and hourly inputs, with a configurable prediction threshold.

The document describes the method and lists a BTC/USDT futures backtest configuration covering roughly one month, but gives no performance measures or details about training data, validation, or how the model weights were obtained. In the supplied code, the network is implemented through fixed weights rather than an apparent training process, so the prose’s description of training is not substantiated by the example. The approach may trade infrequently and depends on sound model validation; data requirements, network design, and threshold choice are noted as concerns.

Key ideas

• Predictions from two timeframes must agree in direction before a trade is opened.
• A disagreement between timeframe signals triggers position closure.
• The example uses daily and hourly inputs with a configurable threshold.
• The published code contains fixed network weights and does not show model training.
• The backtest configuration has no reported performance results, and validation details are absent.

Tags

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