A Single-Layer Perceptron for Combining Lagged RSI Values
Summary
This example explains a simple single-layer perceptron that combines several time-lagged readings of a smoothed RSI. Each selected reading is multiplied by a parameter-specific weight, and the weighted values are added to produce a raw score. The inputs can be changed to other indicators or price data, so the example illustrates a general way to aggregate a small set of observations over time.
The score is not activated into a binary signal in the example. The text suggests defining thresholds to turn it into buy or sell conditions, and says the input lags and weights can be adjusted, potentially with walk-forward optimization. It provides no backtest results or evidence that the suggested RSI thresholds are profitable. The method is an illustrative starting point rather than a trained neural network: it does not describe a learning procedure for estimating weights, and any threshold or parameter selection would need careful out-of-sample evaluation.
Key ideas
- The example combines multiple lagged readings of a smoothed RSI into one weighted score.
- The same structure can accept other indicator or price inputs.
- A raw score can be converted into trading conditions by applying chosen activation thresholds.
- Weights, input lags, and thresholds require evaluation, and the document reports no performance results.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.