<Final Year Project: AI Planning – Robotic Surgery Assistance>

PDDL Modeling: I created various PDDL domains and problem files to accurately represent different surgical procedures, each with unique levels of complexity.
Custom Solver Development: Using Java, I developed a custom solver that incorporates A* algorithm principles to generate surgical plans based on PDDL representations.
Testing and Evaluation: To assess the planner’s performance, I benchmarked it against the Fast Downward and LAMA planners. While my custom solver yielded promising structural results, its performance remained partially inconclusive due to development constraints.

Further dissertation information available on request.

<Sentiment Analysis Tool>

Scope: Leveraging the New York Times comments dataset to analyse public sentiment on a wide array of topics discussed in the media.
Primary Goal: To assess how readers react to various issues, ranging from politics to culture, using sentiment analysis and text categorization techniques.
Approach: To manage the computationally demanding tasks of processing large volumes of text data, I am incorporating the resource-intensive calculations in C for high-performance text parsing and efficient handling of the data.
Additionally, I am building a user-friendly interface with Python using the Streamlit library, which will allow for interaction with and visualization of the data through a dynamic dashboard.

Enable users to generate a 'smart' workout, after selecting from various input choices.
The input parameters: Sport + Goal.
The output given will be a structured workout split(7-day).
I am approaching the problem as a Machine Learning Classification task.