- Surveyed machine learning papers and explored new ways to facilitate circuit partitioning problems- Implemented and compared Self-Organizing Map, K-means clustering, and Neural Network approaches- Focused on using Graph Convolution Network + Fully Connected Layers to do unsupervised learning- Generalized a trained model to fit different size of input graph range form hundreds to 100k+ nodesWorked individually on a research project. Through the guidance of my mentor and manager, learnt how to utilize circuit simulator and familiarize with the designing workflow.Feedback includes: doing a good job in self researching, excellent communication, and good problem solving strategies.