Salaar Mir
Computer Science at University of Nottingham
Computer Science graduate from the University of Nottingham with an MSc in Computational Neuroscience, Cognition & AI. Interested in backend software engineering and applied machine learning, with experience building Python-based systems, data pipelines, and ML-driven applications. Particularly motivated by roles that combine solid engineering fundamentals with real-world impact.
Work by Salaar Mir
- Spiking Neural Networks with FORCE Learning
This project explored biologically inspired learning by implementing recurrent spiking neural networks trained with FORCE learning. I studied how temporal patterns are learned and maintained, analysing the stability, robustness, and generalisation of network dynamics under noise and parameter…
- CNN-Based Visual Navigation System
I developed a convolutional neural network–based navigation system that mapped camera inputs to control actions for an autonomous vehicle. The project focused on training and evaluating models under real-time constraints, analysing latency–accuracy trade-offs and the impact of model complexity in…
- Bayesian Reinforcement Learning for Cognitive Flexibility
This project investigated how agents adapt their behaviour under uncertainty using Bayesian reinforcement learning. I designed and implemented simulation pipelines in Python to compare Bayesian and non-Bayesian agents across changing task conditions, analysing learning dynamics and strategy…