Flight delays are to be predicted by INRTU students within «Project Activities» discipline

Students of the School of Aircraft Construction, Mechanical Engineering and Transport are engaged in predicting flight delays using machine learning methods. The curator is postgraduate student Alexander Gruzdev

Predictive modeling is conducted within «Project Activities» discipline which is included in the schedule of all 3-4 year students this academic year. Classes are held every Monday.

The relevance of the project is determined by the fact that the introduction of artificial intelligence and machine learning methods is the main direction in engineering. A large international airport is able to accept up to 15 million passengers annually, to serve up to 160 thousand landing operations. In such conditions, there is a need to distribute air and ground traffic and collect statistics.

Project leader Alexander Gruzdev sees building predictive models as a training environment where students will gain additional competencies. The key skill is working with real data, which needs to be adjusted.

«We started studying with programming fundamentals and then we analyzed the «Titanic» dataset. The database contains information about passengers, classes of service. The students, using the methods of predictive analytics, mastered the application of predictive models», said Alexander Gruzdev about the studying process.

This semester, senior students will build a basic model based on publicly available airport data, such as weather, aircraft type, and load factor. By solving the problem of binary classification, students will be able to tell if a flight will be delayed or not.

According to Alexander Gruzdev, the project has great prospects for development and will be offered to the management of Irkutsk International Airport:

«By expanding the capabilities of the service, we will be able to inform passengers that a flight has been cancelled or rescheduled for various reasons to another time».

A third-year student, Vadim Ponomarev, participates in the project. At the university, he majored in automation of technological processes and production and worked as a technician in the department of information security and document management.

«During the summer practice, I was building and programming CNC machines and developing parts. Now we are learning how to build basic machine learning models, such as linear regression, using artificially generated data. These skills will be useful for me», shared his opinion Vadim Ponomarev.