LSTM Functional Model for Predicting CSI 300 Returns

Summary

This preliminary example applies a functional LSTM network to Chinese equity data, using the CSI 300 as its prediction target. The sequence input contains six market features—closing, opening, high and low prices, amount, and volume—across 30 time steps. The LSTM representation is combined with an auxiliary label derived from the close price and fed into a dense layer; the stated output is a five-period future return.

The document points to a strategy example but includes no model details beyond this outline, validation results, benchmark comparison, or performance evidence. It does not explain the label's role, how training and testing are separated, or how predictions become trades. The page presents an initial exploration, so the architecture should be treated as a sketch rather than a demonstrated forecasting method.

Key ideas

• The example uses 30 time steps of six price and trading activity features as LSTM input.
• It combines the LSTM output with an auxiliary label before a dense layer.
• The stated prediction target is a five-period future return for the CSI 300.
• The document provides no validation or trading performance evidence.

Tags

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