ATROPS (Adaptive Transportation Optimization Service)

- Localized Climate Predictions for Proactive Wildfire Risk Mitigation- Developed a Seq2Seq model for predicting the Heat Stress Index (HSI) with high accuracy, enabling proactive responses to climate extremes- Achieved a 95% accuracy rate in 6-hourly HSI forecasts, significantly improving the predictiveness of climate impacts on energy demand and wildfire risks

- Embedded Systems, Radio Communication System

- Performed key techniques of neural networks, including gradient descent and backpropagation- Built upon calculus and linear algebra applications- Projects: Used a pre-trained image classifier to identify dog breeds & coded an image classifier- Python, NumPy, and PyTorch

- Biomedical Signal Processing and Modeling: researched how COVID-19 increases susceptibility to various metabolic syndrome symptoms leading to myocarditis- Visualized data through MATLAB by implementing LDA and SVM algorithms which effectively reduced the dimensionality of our dataset and enhanced the classification accuracy of our ML model- Science writing: researched the header bidding ecosystem and the impact on economics

- Learned classification, regression, neural networks, neural nets for classification, learning features, deep fakes, reinforcement learning, vision and language, ethics, search, and probabilistic models- Developed a deep convolutional neural network in Python using Tensorflow and Keras that optimizes plant disease detection with 97% accuracy