readme.md

Input data

Sample data

Sample data overview

• Input data for sample code after chapter 2
• Use data from Kaggle competition Prudential Life Insurance Assessment as a reference. Data was made artificially to simulate insurance underwriting data. The data construction was simple, so its structure is simpler than real life data.
• Total of training and test data is 10000 lines

Sample data items

Column name	Notes
age
gender
height
weight
product	product type
amount	insurance premium
date	application date
medical_info_a1/a2/a3	medical information - continuous variable
medical_info_b1/b2/b3	medical information - continuous and catergorical variables
medical_info_c1/c2	medical information - continuous and catergorical variables
medical_keyword_1-10	medical information - binary variable
target	target values (binary)

Input data used in chapter 1 (ch01-titanic)

• From Kaggle competition Titanic: Machine Learning from Disaster, save following data (save into folders as follows: ch01-titanic/train.csv, ch01-titanic/test.csv)

Input data used in chapter 3 (ch03)

• Data for explanations on how to combine different tables
• Data for explanations on how to process time series data