PhD in Statistics and Machine Learning

Updated: about 1 month ago
Deadline: 01 May 2021

The Operations Management department at the Amsterdam Business School (University of Amsterdam) invites applications for a PhD position in Statistics with a computer science orientation. We are looking for candidates with the ambition to work and succeed at the highest international academic level.

What are you going to do?

The research area is Predictive Process Modeling (PPM), which focuses on forecasting potential problems during process execution before they occur. Applications have been developed in a wide range of domains, such as manufacturing, healthcare, networking, and business processes. Increasingly comprehensive data collection provides more process visibility. Advances in machine learning make use of this increase in detail and frequency of data. Data-driven techniques in this area can be used to improve process quality control by forecasting and monitoring potential process problems. Prediction methods include regression techniques, support vector machines, decision trees, random forest, elastic nets, neural networks, and gradient boosting.

Predictive monitoring starts with defining an (unwanted) process outcome. Subsequently, a model is specified for the process. The parameters of the model are then estimated using the available data. When monitoring commences, the estimated parameters are used to generate process predictions. The probability of the defined process outcome is then calculated. If the probability exceeds a predetermined threshold, the procedure signals. The parameters can then be re-estimated and the monitoring continues. Note that machine learning techniques require less formal modeling, but more data and computing power to perform predictions. Furthermore, as with Statistical Process Monitoring, the threshold to signal will determine the expected false alarm rate. The performance of a PPM procedure can be evaluated using the precision and recall metrics.

The PhD student will work in close collaboration with the supervisory team and other faculty on tasks that include:

  • understanding statistical modeling and applying machine learning techniques;
  • identifying novel research questions based on real-world phenomena or extant theory;
  • presenting research findings at international conferences;
  • writing up findings for publication in international journals;
  • basic programming skills;
  • attending classes and seminars (including those offered at other universities) to further develop thinking and research skills;
  • participating in and contributing to departmental research functions (PhD Day, research seminars, research meetings);
  • teaching-related activities (to a limited degree), including undergraduate tutorials and the supervision of MSc and/or BSc thesis projects.

View or Apply

Similar Positions