From the Kaggle repository of datasets, this one contains 6 features concerning house characteristics of 924 rows (740 for training and 186 for testing). The task is to predict the price of the house, hence the proble will be formulated as regression task*.
Data Source: https://www.kaggle.com/datasets/thomasnibb/amsterdam-house-price-prediction
Here are the features and their data types:
- Address:* text.
- Zip:* text.
- Price:* int
- Area:* int
- Room:* int.
- Lon:* float
- Lat:* float
During feature engineering stage a new statistically signifficant feature was created: Manhattan Distance. Its values were calculated using Lon and Lat.
In the end there is a Linear Regression model, which uses Area and Manhattan Distance features.