<ul><li>Stock market indices are essential for measuring market dynamics, but accurate price prediction has been challenging.</li><li>Existing methods treat indices as isolated time series and use regression tasks, which overlooks the interdependencies among constituent stocks.</li><li>The proposed Cubic framework addresses these limitations by introducing fusion in the latent space, binary encoding classification, and confidence-guided prediction and trading.</li><li>Experiments show that Cubic outperforms existing methods in stock index prediction, achieving better forecasting accuracy and trading profitability.</li></ul>

Why Regression? Binary Encoding Classification Brings Confidence to Stock Market Index Price Prediction

Discover more