"Tell me about your dream house and I'll tell you what it costs." ๐
In this Supervised Machine Learning project I analyze a dataset of housing data and train machine learning algorithms to predict house prices. It is a regression approach, meaning that the predicted labels (house prices) are on a continuous scale.
This is a learning project, in which I practice machine learning techniques such as feature selection, grid search with cross validation and pipelines consisting of preprocessing and modeling steps. I test a variety of models with different architectures to find the best performing ones.
I use the best model to submit a prediction to the associated indefinitely running Kaggle challenge.
The project is separated into 4 working notebooks:
- 01: Data Preparation and Exploration
- 02: Baseline Model
- 03: Feature Selection
- 04: Model Training and Selection
The dataset originates from the challenge's website and can be downloaded from the "Data" tab.
A detailed description of the features can be found at the end of this notebook
and stems from the data_description.txt
file from Kaggle.
This Machine Learning project was carried out in the context of a 4.5 month-long Data Science bootcamp with WBS Coding School. Many thanks to WBS Coding School and to my instructors for the guidance.