Real-estate market is an area, where we can find huge amount of data. On the real-estate web pages there are ten thousands of flats, houses and other realties. Because of the huge amount of the flats, the information stored in this database is incomprehensible for the human mind.
Data mining is a young and emerging field of computer science. It deals with discovering humanly understandable patterns in large data sets. The purpose of data mining is to extract the hidden knowledge from these huge datasets.
The goal of my thesis is to analyse the real estate market of Budapest, show the typical market phenomenon on the real-estate market, and develop a price predicting method, which can effectively estimate the price of an unknown-priced flat based on the other features.
In my thesis I used two different models to make predictions: the K-nearest neighbour method, and the linear regression. Since they are very different solutions for the same problem, I also compared them from different point of views.
During the data mining project, I used CRISP-DM (Cross Industry Standard Process for Data Mining) methodology. This thesis contains the whole documentation of the data mining project, from the business understanding to the presentation of accomplishment.