November 2020 Summaries

Machine learning engineers spend a significant amount of time preparing data rather than analyzing it. Data preprocessing is a crucial bottleneck in ML Python development process, and storing large datasets can be challenging. Activeloop's open-source package Hub offers an efficient solution for managing unstructured data like satellite images, allowing users to store and access the data easily with minimal effort and costs. This tool has helped my team save time and resources during a project focused on modeling economic well-being using satellite imagery and ground truth survey data.