Financial Signal Processing
Currently, Professor Constantinides leads the Financial Signal Processing Laboratory and his latest research is focused on the following topics:
- Short Term Prediction and Long Term Forecasting of financial time series using unique advanced Digital Signal Processing and Machine Learning algorithms, including also advanced stochastic techniques through tailored Kalman Filters, Particle Filters and others.
- Moving Horizon Portfolio Optimization based on Adaptive Constrained Optimization Algorithms. Targets can include typically Mean-Variance, including Asymmetric Variance, Risk Minimization, Optimal Hedging.
- Unique Very Low Latency Financial Data Smoothing and Modelling
- New algorithms from unique areas of Surrogate Time Series Models, Endomorphic Models, Instantaneous Frequencies/Amplitudes/Data Bandwidths, Financial Data Dominant Component Models and Prediction.