: DAE-GRUSA stands for DeepAutoEncoder-Gated Recurrent Unit Sentiment Analysis . It combines a deep autoencoder (DAE) to denoise historical stock data with Gated Recurrent Units (GRU) to process temporal patterns alongside news sentiment scores.
: The model is designed to improve the accuracy of stock close price predictions by mitigating market volatility noise and incorporating qualitative data from financial news. Download GrUSA pdf
: A version of the article is hosted on PMC for free viewing and download. Download GrUSA pdf