Machine Learning Directed Study: Report 2

This document discusses a machine learning directed study project focused on gathering and processing data related to debris in space. The author initially obtained 3D models from GrabCAD to simulate data collection. The models were processed using Blender, resulting in 108 unique parts. The author implemented an algorithm to calculate essential debris properties, such as moments of inertia, using a paper by Eberly. The dataset was normalized based on volume for efficient processing and analysis. A summary of the dataset with scaling is provided, highlighting the properties of the parts. Principal component analysis (PCA) was performed to identify the properties capturing the most variation in the data. The author suggests that the dataset needs to be expanded with more diverse debris and additional properties. The document concludes with references to relevant sources.

This content was originally posted on my projects website here. The above summary was made by the Kagi Summarizer