Develop a machine learning model based on the features availability




I am trying to develop a classification model. This is to predict that the given task will be completed in the given time or not. When I have gone through the data there are many null values and the task is also contains different phases. So I am trying to develop a predictive model depends on the features availability . For example I the the task has 3 features the model should be run with 3 features only, if it has 5 features it should be run with 5 features. Please let me know that how we can achieve it and of there is any sample code that will be appreciated.

Thank you,


Hi @ursraja4u,

Since it is a classification problem, you can start with simple logistic regression. As you have mentioned there are null values in some features, you can impute those missing values or remove the feature based on the percentage of missing values in it. If the missing values in a feature are more than say 40-50%, you can drop that feature otherwise you can impute it with different techniques(like mean, median, mode and many more).

You can refer to the Loan Prediction Practice Problem (using Python) course on the trainings portal of analytics vidhya to learn how to solve a classification problem. In this course, we have explained various algorithms to solve such problems. Here is the link of the course: