Christian James Pollitt PhD Researcher | Bayesian Modeling | Signal Intelligence | Generative Systems
Set the world right; Do what's best - as above, so below.
Founder @ Dreamwrapped, MI.AI
🧠 Profile
Machine learning researcher focused on probabilistic modeling, adaptive inference, and signal classification.
My work sits at the intersection of Bayesian simulation, generative systems, and sensor fusion, applied to real-world trajectory and signal data.
Exploring Generative Modelling, AI and belief propagation.
Building tools for perception.
🔍 Research
Trajectory simulation & classification (OU models, 6DOF)
Bayesian inference (SMC, RJMCMC, adaptive MCMC)
Signal processing (EEG)
Simulations and Automatic tagging
Probabilistic pipelines for noisy, high-dimensional data
🔬 Recent Projects
Novel Algorithm Design
Variational Inference and Generative Modelling.
OU Behavior Engine
Classification and Outlier Detection using Parameter Inference.
Image Segmentation and Automatic Labelling
Based on methods to automatically label large data.
EEG Signal Decoding (Legacy)
Preprocessing, segmentation, and classification of EEG for biometric and neurocognitive research.
🛠 Tech Stack Languages: Python, MATLAB, LaTeX Libraries: NumPy, PyTorch, SciPy, pandas, seaborn, scikit-learn Inference: MCMC, SMC, RJMCMC, Kalman, OU, GP Domains: AIS/radar sensor fusion, time series, EEG, simulation, Bioinformatics, Neuroscience, Systems Biology Tools: Git, Docker, Jupyter, VS Code
🎓 Academic History PhD, Machine Learning & Signal Modeling University of Liverpool, 2021–2025 University of Manchester, Bsc Computer Science & Neuroscience (HCI)
Researching adaptive Bayesian models for behavioral inference across fused maritime and aerial sensors. Emphasis on generative modeling, stochastic simulation, and sequential inference.
📫 Contact 📧 christianjamespollitt@gmail.com
🧠 GitHub: @ChristianJP
🇬🇧 UK-based. Open to research & technical collaborations internationally.