Training Separate Support Vector Machines for Buy and Sell Signals

Summary

The document outlines an example trading system that uses two support vector machines, one for buy signals and one for sell signals. It initializes seven technical indicators, passes their handles as model inputs, and derives training inputs and target outputs from historical price data. After training, the models generate trade signals that trigger orders with manually specified stop-loss and take-profit levels.

This is an implementation overview rather than a documented trading study. It gives no performance results, validation procedure, details about the target-label design, or safeguards against overfitting and data leakage. The indicator set and risk parameters are left open for experimentation, so the description does not establish that the signals are profitable or transferable across markets.

Key ideas

• Separate SVM models are created for buy and sell signals.
• Indicator handles are collected into an input array for the models.
• Historical prices are used to construct training inputs and outputs.
• Trades use manually chosen stop-loss and take-profit levels.
• The example provides no evidence of out-of-sample performance.

Tags

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