Toktam Khatibi

<p>Designed machine learning models for high-stakes prediction tasks (e.g. clinical risk prediction), achieving up to 92% accuracy and AUC 0.92, improving decision reliability over traditional statistical baselines<br>
Developed end-to-end ML pipelines (data ingestion → feature engineering → modelling → evaluation), enabling reproducible experimentation and scalable deployment across multiple projects<br>
Built deep learning and multimodal models (CNNs, GNNs, self-supervised learning), improving classification accuracy and robustness for complex datasets (e.g. imaging, structured clinical data)<br>
Applied reinforcement learning and optimisation techniques to sequential decision-making problems, improving system performance in dynamic environments (e.g. scheduling, resource allocation)<br>
Designed anomaly detection and predictive modelling systems for large-scale datasets, supporting early risk identification and decision support<br>
Collaborated with domain experts, engineers, and stakeholders to translate research into deployable solutions aligned with operational constraints</p>

Designed machine learning models for high-stakes prediction tasks (e.g. clinical risk prediction), achieving up to 92% accuracy and AUC 0.92, improving decision reliability over traditional statistical baselines, Developed end-to-end ML pipelines (data ingestion → feature engineering → modelling → evaluation), enabling reproducible experimentation and scalable deployment across multiple projects, Built deep learning and multimodal models (CNNs, GNNs, self-supervised learning), improving classification accuracy and robustness for complex datasets (e.g. imaging, structured clinical data), Applied reinforcement learning and optimisation techniques to sequential decision-making problems, improving system performance in dynamic environments (e.g. scheduling, resource allocation), Designed anomaly detection and predictive modelling systems for large-scale datasets, supporting early risk identification and decision support, Collaborated with domain experts, engineers, and stakeholders to translate research into deployable solutions aligned with operational constraints

<p>Led research programmes in machine learning, deep learning, and AI, producing 50+ peer-reviewed publications across high-impact journals<br>
Developed novel algorithms in areas including self-supervised learning, graph neural networks, and ensemble modelling, improving performance over baseline methods<br>
Supervised large-scale research projects involving multimodal datasets and experimental design, ensuring rigorous evaluation and reproducibility<br>
Delivered advanced AI systems integrating NLP, computer vision, and structured data for real-world applications</p>

Led research programmes in machine learning, deep learning, and AI, producing 50+ peer-reviewed publications across high-impact journals, Developed novel algorithms in areas including self-supervised learning, graph neural networks, and ensemble modelling, improving performance over baseline methods, Supervised large-scale research projects involving multimodal datasets and experimental design, ensuring rigorous evaluation and reproducibility, Delivered advanced AI systems integrating NLP, computer vision, and structured data for real-world applications