MLA 003 Storage: HDF, Pickle, Postgres
May 24, 2018
Click to Play Episode

Comparison of different data storage options when working with your ML models. You'll ingest your dataset (CSV, TSV, JSON API, ..) via Pandas. Then you munge the data / Pandas, convert to Numpy & send to your model. Model saves intermediate steps as Numpy->HDF. Publish results of your model's predictions (whether for use in app/website, or for researchers, etc) in a SQL database or CSV.