This submission was made for the Internshala Machine Learning Challenge. The challenge involves predicting the ability of newly recruited and trained business agents to win sales orders. For a complete description please see the pdf file in the repo. This was a binary classification problem. The train dataset had string, numerical variables and dates and missing variables. The dates were parsed as datetime objects, the missing values were filled, the string variables were encoded as numbers and the dataset scaled. A part of the dataset was set aside as validation set. Logistic Regression, Bagging with Decision Tree and Logistic Regression, Random Forest were tried to predict the test data. The final result was obtained after obtaining the optimal parameters by hyperparmeter tuning of the Random Forest model
nehalagrawal186 / internshala_ml_repo1 Goto Github PK
View Code? Open in Web Editor NEWThis project forked from me-nobody/internshala_ml_repo1