Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Prediction & valuation of real estate prices with machine & deep learning techniques

Sarapanis Ioannis

Full record


URI: http://purl.tuc.gr/dl/dias/D3E1AF62-E89D-4AA1-9AE1-5FD7BF1B7E7D
Year 2023
Type of Item Master Thesis
License
Details
Bibliographic Citation Ioannis Sarapanis, "Prediction & valuation of real estate prices with machine & deep learning techniques", Master Thesis, School of Production Engineering and Management, Technical University of Crete, Hellenic Army Academy, Chania, Greece, 2023 https://doi.org/10.26233/heallink.tuc.96312
Appears in Collections

Summary

In recent years, an increased interest in trying to estimate real estate prices, with various developed Machine & Deep Learning techniques, has increasingly developed in the research community. In this diploma thesis using computer science and specifically the method of artificial intelligence and machine learning, the process of designing, analyzing and processing mathematical models with the estimation of the value of real estate and finally in the evaluation and selection of the most reliable model. More specifically, different machine learning models are implemented, which are trained based on the data. The majority of these models rely on regression techniques and algorithms, such as multiple linear regression, which is the basis for the development of more complex and efficient regression techniques, such as Ridge regression, Lasso, and Gradient regression Boosting. Also, models based on Decision Trees, such as Random Forests and other models such as Neural Networks, were implemented. Subsequently, the results of their application to the training and control data are presented and evaluated. For this were used data from famous online platforms in Greece such as xe.gr and spiti360.gr, with the help of which, a relevant picture of the real estate market is given. The results we obtained from the internet platforms and the methods we applied are then analyzed and their accuracy and appropriateness for the present project are evaluated. Finally, some comparisons of our results with results of corresponding researches are presented, the points that can be improved in the methodology are identified which was followed and some future research perspectives for the next studies are proposed based on them.

Available Files

Services

Statistics