Improving parking availability data by utilizing machine learning
Kuusrainen, Iitu (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202504156559
https://urn.fi/URN:NBN:fi:amk-202504156559
Tiivistelmä
The current parking availability data in the city of Turku is incomplete due to various parking permits, such as disabled parking permits and resident permits. The aim of this thesis was to improve parking availability data. This was achieved by combining two different datasets and comparing two different machine learning models. The work was commissioned by the City of Turku.
The datasets were combined using the Python programming language, utilizing the Pandas and GeoPandas libraries. The result of the combination was a new dataset that considers data from both sources and therefore provides more accurate parking data. After the combination, machine learning models could be compared. In this work, an iterative self-learning model and a prediction coefficient model were compared in three different categories.
The result of the comparison was the selection of the iterative self-learning model, specifically the AutoML technique. This work is intended to serve as a basis for building a machine learning model.
The datasets were combined using the Python programming language, utilizing the Pandas and GeoPandas libraries. The result of the combination was a new dataset that considers data from both sources and therefore provides more accurate parking data. After the combination, machine learning models could be compared. In this work, an iterative self-learning model and a prediction coefficient model were compared in three different categories.
The result of the comparison was the selection of the iterative self-learning model, specifically the AutoML technique. This work is intended to serve as a basis for building a machine learning model.