Honestly? Just pick the dataset that doesn’t make you wanna scream.
Jk. But for real, I’ve had luck with clustering similar datasets first, then ranking within clusters. Less overwhelming.
Also, Kaggle kernels often have EDA notebooks—steal their ideas!
Jk. But for real, I’ve had luck with clustering similar datasets first, then ranking within clusters. Less overwhelming.
Also, Kaggle kernels often have EDA notebooks—steal their ideas!
