Abstract
The emergence of Industry 4.0 is making production systems more flexible and also more dynamic. In these settings, schedules often need to be adapted in real-time by dispatching rules. Although substantial progress was made until the '90s, the performance of these rules is still rather limited. The machine learning literature is developing a variety of methods to improve them. However, the resulting rules are difficult to interpret and do not generalise well for a wide range of settings. This paper is the first major attempt at combining machine learning with domain problem reasoning for scheduling. The idea consists of using the insights obtained with the latter to guide the empirical search of the former. We hypothesise that this guided empirical learning process should result in effective and interpretable dispatching rules that generalise well to different scenarios. We test our approach in the classical dynamic job shop scheduling problem minimising tardiness, one of the most well-studied scheduling problems. The simulation experiments include a wide spectrum of scenarios for the first time, from highly loose to tight due dates and from low utilisation conditions to severely congested shops. Nonetheless, results show that our approach can find new state-of-the-art rules, which significantly outperform the existing literature in the vast majority of settings. Overall, the average improvement over the best combination of benchmark rules is 19%. Moreover, the rules are compact, interpretable, and generalise well to extreme, unseen scenarios. Therefore, we believe that this methodology could be a new paradigm for applying machine learning to dynamic optimisation problems.

Abstract
This paper explores robust unconditional and conditional nonparametric approaches to support performance evaluation in problematic samples. Real-world assessments often face critical problems regarding available data, as samples may be relatively small, with high variability in the magnitude of the observed indicators and contextual conditions. This paper explores the possibility of mitigating the impact of potential outlier observations and variability in small samples using a robust nonparametric approach. This approach has the advantage of avoiding unnecessary loss of relevant information, retaining all the decision-making units of the original sample. We devote particular attention to identifying peers and targets in the robust nonparametric approach to guide improvements for underperforming units. The results are compared with a traditional deterministic approach to highlight the proposed method's benefits for problematic samples. This framework's applicability in internal benchmarking studies is illustrated with a case study within the wastewater treatment industry in Portugal.

Abstract

Abstract
Problem definition: Marketplace platforms such as Amazon or Farfetch provide a convenient meeting point between customers and suppliers and have become an important element of e-commerce. This sales channel is particularly interesting for suppliers that sell seasonal goods under a tight time frame because they provide expanded reach to potential customers even though it entails lower margins. In this dyadic relationship, a supplier needs to optimize when to share inventory with the platform, and the platform needs to set the right commission structure during the season. Academic/practical relevance: We characterize supplier participation into the platform in a dynamic setting and link it to inventory levels, demand rates, time left in the season, and commission structure. This directly drives the commission structure decision made by the platform. We, thus, provide a framework to evaluate platform commission fee policies, taking into account supplier responses. Methodology: We use an optimal control framework with limited inventory supply and a stochastic demand process. We study the conditions under which the supplier accepts participation and use the platform as a sales channel. We also study the optimal commission structure that the platform should employ and the supplier procurement response. Results: We find that suppliers only participate if inventory is high relative to the time left to sell the items. As a result, the platform can only offer limited supply at the beginning of the season. Given this behavior, we find that the platform and the system are always better off with flexible pricing via fully dynamic commissions, which hurts the supplier the most (better off with less flexible commission fees). Interestingly, when the inventory decision is contingent on the platform pricing policy, the platform often finds it beneficial to commit to a static fee to incentivize the supplier to stock up, highlighting that inability to commit to fixed commissions may destroy value through double marginalization effects. Managerial implications: Our work suggests that short-term profit for the platform is maximized with fully dynamic commission fees at the expense of supplier profit. If inventory is endogenous, suppliers can retaliate by reducing their commitment at the start of the season. Despite the increased revenue obtained with the fully dynamic commission fee, the lost sales from the inventory drop incentivize the platform to opt for supplier-friendly commission fees, which are better for long-term-profit.

Abstract
Problem definition: Online retailers are on a consistent drive to increase on-time delivery and reduce customer lead time. However, in reality, an increasing share of consumers places orders early. Academic/practical relevance: Such advance demand information can be deployed strategically to reduce costs and improve the customer service experience. This requires inventory and allocation policies that make optimal use of this information and that induce consumers to place their orders early. An increasing number of online retailers not only offer customers a choice of lead time but also, actively back-order missing items from a consumer basket. Methodology: We develop new allocation policies that commit to a customer order upon arrival of the order rather than at the moment the order is due. We provide analytical results for the performance of these allocation policies and evaluate their behavior with real data from a large food retailer. Results: Our policy leads to a higher fill rate at the expense of a slight increase in average delay. The analysis based on real-life data suggests a sizeable impact that should impact current best practices in online retail. Managerial implications: With the changing landscape in online retail, customers increasingly place baskets of orders that they would like to receive at a planned and confirmed moment in time. Especially in grocery, this has grown fast. This fundamentally changes the strategic management of inventory. We demonstrate that online retailers should commit early to customer orders to enhance the customer service experience and eventually, to also create opportunities for reducing the cost of operations. Superscript/Subscript Available

Abstract
This paper aims to identify and analyze the factors that influence the decision of Mato Grosso’s farmers to produce soybean using the Analytic Hierarchy Process (AHP). We found evidence that decisionmaking of soybean production is related to rural production aspects such as climate, financing, cost of inputs, and soil quality rather than marketing and logistics. The novelty of this paper is the empirical analysis of the decision-making in agricultural production using AHP. The decision model was created and tested considering 21 farmers and 19 experts linked to the soybean production. Three different scenarios were considered: farmers’ view, experts’ view, and combined view. Our findings indicate that farmers and experts agree with rural aspects are predominant in the decision to plant soybean. Moreover, logistics have been used as an important flag of soybean competitiveness on international trade by soybean stakeholders in Brazil. However, our results show that logistics impact in the soybean decision-making process is low. Due to data limitation access, this study focuses only on Mato Grosso. However, this study has an exploratory character and presents empirical results that may help to understand soybean production over the country.