Machine Learning

I am passionate about machine learning and related fields.


My interests are:

  • Preference Learning
  • Meta-Learning
  • Intelligent Decision Making / Reinforcement Learning
  • Computer Vision

The ML methods I am currently interested are:

  • Probabilistic Ranking Models
  • (Deep) Neural Networks
  • Online Algorithms (for Large Scale Applications)
  • Genetic Algorithms
  • MDS techniques

Edge as well as cloud computing are the major platforms for realizing ML based applications. As a software engineer I am therefore also interested in those exciting technologies.

Ranking of Paired Data

For the purpose of learning from and the prediction of preferences I worked on a new supervised learning setting called dyad ranking (see thesis).


The following problems can be tackled with it:

  • Classification (binary, multi-class, multi-label)
  • Similarity Learning
  • Label Ranking

Related problem settings are:

  • Learning to Rank for Information Retrieval
  • Collaborative Filtering & Ranking
  • Dyadic Prediction
  • Zero-shot Learning

Recent publications:

  • Preference-based Reinforcement Learning using Dyad Ranking [paper]
  • Dyad ranking using Plackett–Luce models based on joint feature representations [paper]
  • A Latent-Feature Plackett-Luce Model for Dyad Ranking Completion [paper][slides]
  • Plackett-Luce Networks for Dyad Ranking [paper][slides][code]
  • Preference-Based Meta-Learning using Dyad Ranking: Recommending Algorithms in Cold-Start Situations [paper]
  • Dyad Ranking using a Bilinear Plackett-Luce Model [paper]
  • Nameling Discovery Challenge - Collaborative Neighborhoods [paper]