Jul 2026
CLaC@FinMMEval 2026 Task 3: Sentiment-Augmented Deep Reinforcement Learning for Active Trading - An Alpha-Reward Approach
This paper presents the system for Task 3 of the CLEF 2026 FinMMEval Lab, which requires daily long, flat, or short trading decisions for Bitcoin and Tesla using news and historical market data, and achieves the strongest overall performance.
Andrei Neagu, Eeham Khan, Leila Kosseim
· arXiv.org · 0 citations