why data driven machine learning is so hard | tips for improving your datasets
why is data driven machine learning so hard? even in the age of big data, there are still some major issues that arise when building machine learning algorithms related to the way we curate and annotate our data.
today i'm gonna talk some basics on how i think about data while training new algorithms, and tips on getting good results when your data isn't the best.
#fancyfueko #machinelearning #datasets
Introduction (0:00)
My Algorithm Development Process (1:51)
Train vs. Test vs. Validation Data (2:05)
Model Size, Dataset Size and Dataset Diversity (2:32)
The Issue with Dataset Bias (3:56)
Ways to Combat Dataset Size and Diversity Issues (5:28)
Conclusion (6:57)
Related Videos + Subscribe (7:07)
If you like this video, subscribe to my channel:
/ @fancyfueko
Follow me on IG: @fancyfueko
/ fancyfueko