Exploratory Data Analysis (EDA) in Machine Learning - ML 075
Adventures in Machine Learning
00:00
Is Your Data Warehouse a Data Warehouse?
If there's leakage, your fit will be crazy good and deceptively so. If you want to look for correlations, this can indicate whether features are valuable. What other visualizations or EDA tools are in your toolkit that you use for most projects? So you mentioned scatterplot, ANOVA, correlations, and then subject metanollege about leakage. What else? One of the things that I do is check with back-end engineering, whoever's generating that data,. And I ask them, when is that data created?
Transcript
Play full episode
