AI Developments in FinTech
Here are two completed projects as examples:
1. Proof-of-Concept: AI-Powered Stock Selection Using Explainable AI
- Objective: Develop data-driven rules for stock picking with high potential for price increases or decreases.
- Approach:
- Analyze quarterly corporate financials using explainable AI.
- Identify data-based structures guiding stock selection through explainable AI.
- Generate clear, interpretable rules for human evaluation and decision-making.
- Benefits:
- High success rates in stock selection with transparent decision foundations.
- Enables investors to make informed assessments of investment options based on data-driven insights.
- Automates the stock-picking process for recurring decisions.
2. DSS with Dashboard for AI-Powered Futures Trading
- Objective: Develop a robust decision support system for trading energy front-end contracts.
- Approach:
- Combine and integrate four data sources, including market prices, web analytics data, and commodity prices.
- Use bionic swarm intelligence for variable selection.
- Implement a self-learning system that updates daily to avoid concept drift.
- Integrate a mechanism to block predictions when external factors (e.g., regulatory changes) strongly influence market conditions.
- Results:
- A live test over 260 trading days demonstrated an 80% return and 60% prediction accuracy.
- Server-side architecture and a user-friendly dashboard provide daily forecasts.
- Risk assessment through simulated position management and integration of customer-relevant financial metrics.
- Benefits:
- DSS to support traders with clear risk assessments.
- Efficient management of complex trading parameters and external influences.
- Increased trading returns through data-driven optimization.