F²Agent：面向多模态交易的金融智能体融合范式
原标题：F$^2$Agent: Financial Fusion of Agentic Intelligence for Multimodal Trading
AI 摘要
arXiv 论文提出 F²Agent，一种用于多模态金融交易的新型智能体范式。它通过分层专业智能体提取模态特定信号，并引入模态感知自适应融合机制和噪声鲁棒一致性正则化，以捕捉跨模态依赖并生成抗噪交易信号。在六种股票和加密货币资产上的实验表明，F²Agent 在多项交易指标上优于 16 个基线，平均年化收益率相对提升超过 20%，在 GOOG 和 TSLA 上分别实现 120.48% 和 148.41% 的回报。
正文节选
Computer Science > Multiagent Systems Title:F$^2$Agent: Financial Fusion of Agentic Intelligence for Multimodal Trading View PDF HTML (experimental) Abstract:With increasingly diverse and heterogeneous information sources, effectively leveraging multimodal data is becoming pivotal for high-quality financial trading. Although recent advancements in Large Language Model (LLM)-based agents have enabled the ingestion of multimodal inputs, existing methods fail to capture nuanced cross-mo