Estimating lead time using Markov Chains

The goal of this thesis is to create a prediction model that can estimate the lead time of LFL (Load Forming Logic), a software product of Vanderlande Industries B.V. The research started by investigating order data, such as weight and volume, and continued with a series of regression, ensemble and Markov chains models. Out of all the proposed solutions, two models have been chosen: Elastic net regression and Markov chains. The results showed that even though Elastic net regression had better results, it is more sensitive to LFL's configuration parameters and different inputs than Markov chains.