Machine learning algorithms for multi-label classification problems.

Machine learning algorithms for multi-label classification problems.


Classification is a classical and fundamental problem in the field of Machine Learning. In normal classification problems, an object is classified to only one class, where the total number of possible classes can be multiple, i.e., this is a "single-label, multi-class" problem.

However, in some applications it is not enough. e.g. A movie, song or book may have multiple genres. An image may have several objects. To assign multiple classes to such an object, more complex Machine learning algorithms for the "multi-label, multi-class" problem must be implemented.


In this thesis a Machine Learning model for multi-label classification problems is to be developed. Most importantly, the model shall be able to predict after training the label assignment can make statements about the label assignment. Furthermore, it should also be able to suggest possible solutions in case of doubt.


  • Desire and motivation for new technology
  • Ideally experience in
    • Programming (Python, C++...)
    • machine learning