Delivery-Truck-Ontime-Delay-Prediction

Source data was taken from Kaggle

Background

Delivering goods from one state to another has been very popular of using trucks. However, more than 35% cases are considered as late/delay shipment. Even in some case, the delivery supplier managed to got the goods delivered delayed more than 1 month. This is causing anxiousity from the customer that going to use that delivery supplier.

Goals

• Create a model that able to predict whether the delivery will be delayed or on time
• Decide top 10 features that has the highest ratio to determine the delivery status

Step

• Eliminate uncorrelated features
• Make some new features that has stronger correlation to the target variable
• Input some missing data with mode and correlated other features
• Try fit model using :
  • Decision Tree Classifier
  • Logistic Regression
  • Gaussian Naive Bayes
  • Random Forest Classifier
  • Gradient Boosting
  • K Nearest Neighbors

Here is the results of the fitted model:

Search the Most Important Feature regarding to the prediction

Using best fitted model (Gradient Boosting with accuracy of 89.7%), try to find the feature importance. Here is the result :

10 most important features are :

• GpsProvider
• vehicleType
• vehicle_no
• Distance(Km)
• org_state
• supplierID
• org_city
• des_city
• customerID
• des_state

Hope this repo is helpfull and got you some new insights

Cheers ! Thankyou :)