Research

Forecasts support decisions.

You’ll find this statement in many forecasting papers and this statement always intrigues me with many questions: really? you sure? how?

This statement has two main objects: forecasts and decisions. Both are achieved in different ways. For example, one could produce sales forecasts using a very simple naive method or a very sophisticated machine learning algorithm. On the other hand, often, mathematical and statistical models provide decision makers with specific decisions. For example, a specific inventory control provides an optimal number of how much to order.

Both problems have different objectives and purposes, and I am interesting in linking these objects – forecasts and decisions.

Currently, I am working on forecast congruence with Nikolaos Kourentzes. It is a new metric to connect between forecasts and supply chain management [link].

At heart, I enjoy working on statistical forecasting models. With Ivan Svetunkov, I am working on incorporating parameter uncertainty in predictive distributions. With Huijing Chen and Ivan Svetunkov, we are expanding the vector exponential smoothing methodology.

PhD Supervision

I am happy to supervise PhD students in business forecasting and its interface in inventory management and tourism. These are topics that I am interested in:

Demand forecasting and inventory management interface;
Advanced development of statistical vector forecasting models;
Shrinkage estimators for single-source-of-error state space models;
Further development of stability in forecasts and decisions;
Exploring the concept of forecast trustworthiness;
Explainable models and forecasts in tourism settings.

Potential PhD students should send an inquiry to me directly via emails before applying to the school. Please include your arguments on why mitigating parameter uncertainty is important for demand forecasting and its subsequent decisions.

Publications

Pritularga, K. F., Svetunkov, I., & Kourentzes, N. (2023). Shrinkage estimator for exponential smoothing models. International Journal of Forecasting, 39(3), 1351–1365. [link]

Pritularga, K. F., Svetunkov, I., & Kourentzes, N. (2021). Stochastic coherency in forecast reconciliation. International Journal of Production Economics, 240. [link]

Working papers

Pritularga, Kandrika and Kourentzes, Nikolaos, Forecast Congruence: A Quantity to Align Forecasts and Inventory Decisions (January 31, 2024). Management Science Working Paper No. 1., Available at SSRN: [link]

Presentations and conferences

Quarterly Forecasting Forum, Spring 2025, London, United Kingdom

5th Institute of Mathematics and its Application and OR Society, Spring 2025, Bimingham, United Kingdom

44th International Symposium on Forecasting, Dijon, France

The OR Society’s annual conference for 2023, Bath, United Kingdom

Quarterly Forecasting Forum, Winter 2023, Lancaster, United Kingdom

42nd International Symposium on Forecasting, Oxford, United Kingdom

39th International Symposium on Forecasting, Thessaloniki, United Kingdom