5 Comments

Great article. Maybe a typo here: the formula for accuracy is (TP + TN) / (TP + TN + FP + FN)

Expand full comment

Ah, thanks for catching that! I have corrected the formula.

Expand full comment

Wonderful analysis ! The Saad et al posted their data/models in figshare: https://figshare.com/projects/EqPred/146166. Inspired by your analysis, I plan to analyze and figure out the model they used and provide additional analysis of why these ML-based approaches isn't gonna solve prediction/forecasting challenge :)

Expand full comment

Fabulous! Let us know what you find.

Expand full comment

👍🏻🧨dynamite article!

Expand full comment