To determine probabilites for a given classification problem

r
machine_learning
data_wrangling
random_forest
data_science

#1

i have the medium based company with web based business.
whenever a customer will login in to website he will be assigned by some score, based on which page he enters he will get some score, meanwhile some important features are present in portal and whenever he use the feature user will get high weight-age . i have 16 plans (paid and free categories) .

My data: paid -0.005% free-99.41% ( till now)
customer id(PK) / paid_free(free-0 or paid- 1)/ plans(plan1 to plan 16)
/action1/./././action146

1.now i want to predict the probability as which customer is going to convert from free to paid.
2.which parameter is going to make the customer to convert him or which parameter is important compared with paid_free
3.need multiple algorithms to compare the accuracy of the model


#2

Hi,

Can you please share the dataset or print the head of your dataset which will help us understanding the question and features better.

You can use the feature_importance in random forest to determine which feature is more important.

For this problem statement, random forest or any other boosting algorithm can be used.