Brief

Researchers have introduced DeSI (Deep Single-Index Fréchet Regression), a new framework designed for predicting outputs in non-Euclidean spaces from high-dimensional inputs. DeSI utilizes a deep neural network to estimate an interpretable index direction, which reveals the relative importance of different inputs. This approach aims to mitigate the curse of dimensionality while maintaining interpretability, offering a contrast to standard deep neural networks. The framework has demonstrated strong predictive performance in simulations and a real-world application. AI