Advanced Consulting Projects
Explore a deep dive into the high-impact analytical work our teams deliver to our industry partners.
Robust Covariance Matrix Estimation (Inspiration-Q)
Developed and rigorously backtested several advanced methods (EWMA, Adaptive Half-Life, Regularized Estimators like Ledoit-Wolf) to forecast covariance matrices in financial time series. This work is crucial for optimizing asset allocation and improving out-of-sample prediction in investment portfolios.
- Focus: Time Series Forecasting, Portfolio Optimization, Quantitative Risk.
- Impact: Comparative analysis and White Paper documenting the most robust methods.
Intraday Liquidity Modeling in Crypto-Assets
Analyzed high-frequency order book and trade data for various crypto-asset pairs to identify intraday liquidity patterns. The project defined a **Liquidity Score** to synthesize metrics (spread, depth, slippage), allowing the client to optimize the timing and execution algorithms (TWAP, VWAP) for large orders.
- Focus: Data Engineering, Liquidity Metrics, Execution Strategy Optimization.
- Impact: Identification of optimal windows for large-volume trading efficiency.
Predictive Risk Detection in Older Adults (YASYT)
Developed a classification and time-series model to identify risk events (like falls, sleep, or mobility changes) in older adults using multisensor IoT data. The project combined anonymized real data with generated synthetic data to train a robust model, exceeding client expectations for distinguishing valid alerts from false alarms.
- Focus: Machine Learning Classification, Time-Series Analysis, IoT Data Engineering.
- Impact: Improved early identification of risk situations and reduction of false alerts for YASYT Robotics.
Affiliated with
