Forecasting Research
I’m an assistant Professor at VIVES University of Applied Sciences (Kortrijk, Belgium). I focus on research on Artificial Intelligence, Sales Forecasting and Leading Indicators.
Look at my blog and use cases and get inspired.
Feel free to contact me
News
Recent posts
- Forecasting with Deep Temporal HierarchiesAs we head for one of the most picturesque student cities of Flanders, Filotas is preparing his research seminar for KU Leuven. Temporal Hierarchies Temporal aggregation is reviewing the data on different frequencies: e.g. monthly, quarterly and yearly. By reviewing the data this way, the data is multiplied: one time series becomes multiple time series.Meer lezen over “Forecasting with Deep Temporal Hierarchies”
- Forecasting case: student performanceWe investigated how students’ performance can be predicted early on in the semester. For this we only used the digital footprint of the students on the learning platform. What about business relevance? These applied research cases require developing a specific model. It will be interesting to investigate the carry-over effects when applying this to businessMeer lezen over “Forecasting case: student performance”
About
Sales
Forecasting
Knowing future demand is essential in optimising your supply chain. Whether it is maximising production, minimising inventory or increasing Return-On-Investment (ROI), sales forecasting can help you.
Markets are becoming more globally connected and are getting very volatile. A better understanding of your customer and the market improves demand planning and customer satisfaction.
Potential impact areas
Robust sales forecasting
A lot of global companies (e.g. Amazon, Bol.com) put in large efforts to improve their sales forecasting with artificial intelligence (A.I.) algorithms. With increasing research & open software, these algorithms are available for other companies.
Leading Indicators
A changing market is very hard to predict. Little models foresee the global 2008 financial crisis or the 2019 covid pandemic in advance. Market intelligence from leading indicators can improve these forecasts. In my research, I automatically detect the most significant market indicators.
Impact on Inventory
The real gain of demand forecasting is in improved customer delivery and satisfaction, while keeping the operational costs (and inventory) to a minimum. My latest research takes this into account and feeds it in the A.I. model.
Practitioners’ references
What the experts say
The [Sagaert] model provided [us] a 15% MAPE improvement in forecast accuracy [..] helping us in reducing our global inventory and saving us thousand of dollars in capital investment and operational expenses.
Director Global Marketing & Business Development
A.I. is moving from the lab to the workplace, with profound implications for business and society.
McKinsey & Company
Collaborate in our research
Are you interested in learning more? Contact me below