Bayesian Neural Networks for uncertainty-aware financial forecasting and probabilistic time-series modeling.
- Implement Bayesian Neural Networks in PyTorch
- Explore variational inference and Monte Carlo methods
- Evaluate predictive uncertainty and calibration
- Analyze financial time-series under noisy conditions
Early-stage research and implementation.
- Data ingestion and preprocessing
- Baseline Bayesian MLP models
- Uncertainty estimation
- Calibration evaluation
- Financial forecasting experiments