Ah yes, the story of my master's ML thesis. I had to select features from multiple papers in order to try and select the best ones with classification results

Ah yes, the story of my master's ML thesis. I had to select features from multiple papers in order to try and select the best ones with classification results to prove it. A few problems included: - Incomplete/unavailable datasets (404 on some copyright pictures) - Features consisted on Math formulas and text descriptions (no code whatsoever) - Classifier names only (which framework did you use? parameter values?) In the end i couldn't contribute as well, got instructions to save my work in a private repo despite being funded by an EU academical scholarship.