Pattern Recognition Tasks with Personalized Federated Learning
This study proposes a personalized federated learning (PFL) approach for pattern recognition tasks, improving model accuracy by adapting to client-specific data while preserving privacy. It demonstrates superior performance in tasks like image classification, speech recognition, and text analysis compared to traditional federated learning methods.
Jan 1, 2025
xFiTRNN: A hybrid self attent linearized phrase structured contextualized transformer based RNN for financial sentence analysis with sentence level explainability
xFiTRNN is a hybrid model for financial sentence analysis, combining self-attention, linearized phrase structures, and a contextualized transformer-based RNN. It improves accuracy in sentiment classification while providing sentence-level explainability. This model is ideal for financial text analysis and decision-making transparency.
Dec 7, 2024