Machine Learning for Intelligent Transportation Systems

Mohammad Bawaneh
Mohammad Bawaneh

For English BSc and Msc students/Angol nyelvű BSc és MSc képzés hallgatói számára szóló téma


Nowadays, the smart city concept is becoming more actual than ever as cities are growing and becoming more and more crowded as a result of urbanization and growth of the world population. The smart city can make intelligent responses to different kinds of needs, including daily livelihood, environmental protection, public safety and city services, industrial and commercial activities. Among the different notable goals of smart cities, construction of intelligent transportation systems could have a significant impact to residents of future cities. Advanced Traffic Management Systems (ATMSs) and Intelligent Transportation Systems (ITSs) integrate information, communication and other technologies and apply them in the field of transportation to build an integrated system of people, roads and vehicles. These systems constitute a large, full-functioning, real-time, accurate and efficient transportation management framework.

In ATMSs and ITSs it is a fundamental challenge to predict the next possible states of traffic with high precision, because this information helps to prevent unlikely events like traffic jams or other anomalies on roads. Therefore the research topic focuses on investigation of prediction methods and techniques, appropriate for optimizing the transport systems, using different sensor information. Machine learning methods for this purpose will be analyzed and utilized. One of the challenges is to deal with different road sections and neighborhoods, where the traffic is correlated, and unexpected events in one can affect the other significantly. For this, classification and clustering algorithms should be developed and validated.

Specifically, you will be involved in one or more of the following Machine Learning tasks (depending on your background knowledge and how many semesters you have before graduation):

  • Data Cleaning and Preprocessing.
  • Traffic Prediction (Using Time Series Prediction Methods).
  • Congestion Detection (Using Time Series Anomaly Detection Methods).
  • Roads Network Clustering (Using Time Series Clustering Methods).

Required skills:

  • Basics of Python programming language.
    • Knowledge of one or more of the following Python libraries is a plus: (Pandas, NumPy, Matplotlib, Tslearn, Keras, TensorFlow, Scikit-learn)
  • Theoretical basics of Data Analysis and Machine Learning.