Suppose you have a dataset with 10 features and 2000 examples for training. You apply Logistic Regression on this data and find out that it is underfitting and hence not giving the desired results. Which of the following steps do you find the most promising to improve the results?
How do we analyse that undefitting data will peform better on another classifier?
Hey @chiragwxN, First step in building models is to use train and test split.
So every time, first load the data and shuffle it. Than make a split using sklearn’s train_test_split. General ratio being 80:20, 80% for training and 20% for validation.
Now you can use any machine learning, deep learning model, just keep it mind that, always use 80% part for training it. Now check score on the remaining 20% data. Now you can judge which model is the best, or which hyperparameters are the best.
Happy Learning 
and mark the doubt resolved if it is 
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.