The aim of the project is to acquire knowledge and practical skills regarding designing and evaluating the performance of neural networks/deep learning-based pattern recognition systems.
The project's emphasis is not only on implementation and experimental evaluation but also on analyzing the results obtained and the appropriate presentation of conclusions.
Project groups can consist of a maximum of two people. The members of the project group must be reported to the teacher. Grades are given individually, and team members share responsibility for the completed project.
The project report must be prepared using LaTeX ub the lncs format. The suggested platform is Overleaf.
The implementation of the algorithms as well as the experimental environment must be realized in Python. Additionally, the implementation must be compatible with the API and tools of PyTorch (neural network architecture) and scikit-learn (evaluation protocol) libraries.
The recommended package for data visualization is matplotlib.
Reports should be uploaded to e-portal.
AI tools are not completely prohibited, but generating whole parts of text or code will be considered plagiarism and thus result in failing the course. For the details regarding the allowed scope of application, please read our department guidelines: https://wit.pwr.edu.pl/en/students/guidelines-for-the-use-of-ai-tools
The documentation should include a report of the AI-based tools used, along with a scope description, in the form of a table, for example:
Submission of the same project to multiple classes without consulting it with the teacher is considered as plagiarism and results in a failing grade.
[M] — mandatory attendance