# 2024 Travelers NESS Statathon — Data 来源：https://www.kaggle.com/competitions/2024-travelers-ness-statathon/data 采集时间：2026-10-19T17:36:08Z ## Data Description The data provided to you and your team consists of variables describing customers that asked for quotes. There are three datasets: at the policy, driver, and vehicle levels. Each row of the policy dataset corresponds to a single policy, which may be associated with multiple drivers and vehicles. Likewise, each row of the driver dataset represents a single driver, and each row of the vehicle dataset represents a single vehicle. It's easy to check that each driver and each vehicle are associated with just one policy. You may assume that all members on a policy live at the same address. Your company was only able to convert a fraction of the policies found in this sample. The policy dataset also has a training and test split variable called split. Note that the conversion indicator (the response variable) is missing for policies in the test split. Your task is to build a model on the training data and apply your model to predict the conversion indicator for each policy in test data. ### Policies.csv policy_id: Unique customer identifier Quote_dt: Date the quote was submitted quoted_amt: Quote amount (US dollars) Prior_carrier_grp: Prior carrier group Cov_package_type: Level of coverage needed discount: Whether or not a discount was applied to the quote amount number_drivers: Number of drivers credit_score: Credit score of primary policy holder num_loaned_veh: Number of vehicles on policy that have a loan associated with them num_owned_veh: Number of owned vehicles on the policy num_leased_veh: Number of leased vehicles on the policy total_number_veh: Total number vehicles on the policy primary_parking: Where car(s) are primarily parked CAT_zone: Catastrophe risk zone Home_policy_ind: Does customer has existing home insurance policy with Peace of Mind zip: US zip code of policy holder state_id: State of policy holder county_name: County of policy holder Agent_cd: Unique agent code (8 digits) split: Train/Test split **convert_ind: Conversion indicator (0=no, 1=yes). This is the response variable** ### Drivers.csv: policy_id: Unique customer identifier gender: Gender of driver age: Age of driver high_education_ind: Higher education indicator safty_rating: Safety rating index of driver living_status: Driver’s living status (levels = ‘own’, ‘rent’, ‘neither’) ### Vehicles.csv: policy_id: Unique customer identifier car_no: Unique car identifier (per policy) ownership_type: Whether the car is loaned, owned or leased color: Vehicle color age: Vehicle age make_model: Make and model of the vehicle # NOTE **Training and Test sets are combined across the files, and need to be separated manually by your group!**