Doubt for Feature Selection

#1

how to selection the best features out of the given features??

#2

and how to apply nueral network for regression,what would be the output_

#3

and how to apply nueral network for regression,what would be the output_size??

#4

Hello Parth,

  1. Feature selection is best done by observing the data. You must know the data you’re working on well before you implement any algorithm on it. Look at the featured you have in your data, a feature like name or something that you can logical sense won’t add to our knowledge of the output would be removed.
    Plus, there’s a method provided by pandas that can be applied on a dataframe which is data.info(), this would print a summary of your data, with how many null values are present in each column. A column with many null values should be removed.

  2. Applying the neural network for regression works in the very same way as it does for classification. Your y_train would now contain the numerical values which are desirable from a set of features and the ‘metrics’ which you’ll define while compiling your model would be R2 Score and not ‘accuracy’.

  3. The output_size would be the number of outputs you want. For example if you want to predict Current House Prices as well as House Prices after 5 years, then the output_size would be 2 and so would be the case in y_train.

I hope this resolves your query.

#5

I hope I was able to resolve your query?

#6

I hope I’ve cleared your doubt. I ask you to please rate your experience here
Your feedback is very important. It helps us improve our platform and hence provide you
the learning experience you deserve.

On the off chance, you still have some questions or not find the answers satisfactory, you may reopen
the doubt.