What if we could not able to find the data pattern or any strong correlation or the data is messed up completely.
How to deal with this?
Hi Anubhav_Gupta,
As in most of the machine learning projects, we usually spend most of the time cleaning our dataset and perform Feature Engineering, which as you know is to extract features from raw data to improve the performance of machine learning algorithms. Even if there is not a direct strong co-relation between the features/attributes, using PCA and other techniques, its possible to combine different features and then find co-relations on those extracted features. so that we can use the processed data in our Algorithms.
I hope it helps.
Thanks!