looking for an apartment in berlin is so annoyingly hard and competitive that it feels like it should be an olympic sport. Everyone knows that finding the perfect apartment isn’t easy, but being well prepared by doing some research on the renting market in Berlin before jumping in and trying your luck, could improve your chances. Recently, I was thinking about moving myself and wanted to get an idea of what I should expect, so I decided to do this project to help me with my apartment hunt!

Project Overview

• Analyzed two different datasets scrapped from Immobilienscout24.de.

• The raw data consisted of rental offers posted on Immobilienscout24.de in Sep 2018, May 2019, Oct 2019, Feb 2020 and Apr 2020.

• Cleaned and joined the two datasets as well as created new features from the description of each listing.

• Conducted an extensive EDA in order to develop insight on the rental market in Berlin.

• Implemented three different Machine learning models that could estimate the total rent price (Warmmiete) of an apartment in berlin.

• Used GridsearchCV to find the best hyperparameters for a Random Forest Regression.

• Serialized the RF model using pickle and created a simple UI.

Results

Evaluation Metric	Multiple linear Regression	Lasso Regression	Random Forest
MAE [€]	194.92	194.92	166.94
RMSE [€]	277.76	277.76	259.80
R² [-]	0.85	0.85	0.87

GitHub repo