.png){kind=icon}
.png){kind=icon}
Return
Event-Driven Financial Market Analyses Using a Multi-Agent Framework
.png){kind=icon}
PRE
AI
Q
V
DOI：10.1109/tbdata.2026.3688995.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
We introduce FinPulse, a sophisticated system that combines machine learning with a multi-agent framework to provide interpretable and event-driven financial intelligence. Utilizing aligned historical data that integrates macroeconomic events with market reactions, FinPulse is capable of predicting asset returns and volatility over various time frames. Specialized agents within the system are equipped to generate model-based causal interpretations, simulate counterfactual scenarios, and convert model outputs into user-friendly narratives. FinPulse exhibits considerable explanatory power and offers actionable insights across a wide range of asset classes. This framework lays the groundwork for future developments aimed at real-time adaptation, expanded event coverage, and enhanced interactive user interfaces.
Keywords:
Multi-agent system
machine learning
time-series forecasting
causal analysis
macroeconomic events
Journal
I
IF:
5.7
Papers:
834
Citations:
3.0K
Organization
Cited Papers