• Simulated specific government policies and costs to find ways to lower global CO2 levels while maintaining high long-term GDP growth using policy-based reinforcement learning in OpenAI’s Gym library.
• Implemented and visualized linear and polynomial regression to predict future CO2 and GDP growth using Pandas, Matplotlib, and scikit-learn.
• Discovered that projected long-term GDP growth can stay positive with the correct CO2 policies; however, the tradeoff is a slight decline in the overall growth rate from 2% to around -1%, indicating a negative growth rate in 100 years.
• Developed an interactive visual representation of the A* pathfinding algorithm, enabling users to visually plot a start and end point on a grid and observe the algorithm's real-time pathfinding process.
• Employed data structures like Priority Queues for maintaining open nodes in the pathfinding process, optimizing the algorithm's performance.
• Developed a supervised learning computer vision model that detected diseased cells using Keras in TensorFlow.
• CNN model was trained on over hundreds of cell images using a validation split to determine what to look for.
• Employed various types of hyperparameter tuning (pooling size, Kernel size, epoch size) through hyperparameter testing.