Abstract

Abstract
Port wine stocks ameliorate during storage, facilitating product differentiation according to age. This induces a trade-off between immediate revenues and further maturation. Varying climate conditions in the limited supply region lead to stochastic purchase prices for wine grapes. Decision makers must integrate recurring purchasing, production, and issuance decisions. Because stocks from different age classes can be blended to create final products, the solution space increases exponentially in the number of age classes. We model the problem of managing port wine inventory as a Markov decision process, considering decay as an additional source of uncertainty. For small problems, we derive general management strategies from the long-run behavior of the optimal policy. Our solution approach for otherwise intractable large problems, therefore, first aggregates age classes to create a tractable problem representation. We then use machine learning to train tree-based decision rules that reproduce the optimal aggregated policy and the enclosed management strategies. The derived rules are scaled back to solve the original problem. Learning from the aggregated optimum outperforms benchmark rules by 21.4% in annual profits (while leaving a 2.8%-gap to an upper bound). For an industry case, we obtain a 17.4%-improvement over current practices. Our research provides distinct strategies for how producers can mitigate climate risks. The purchasing policy dynamically adapts to climate-dependent price fluctuations. Uncertainties are met with lower production of younger products, whereas strategic surpluses of older stocks ensure high production of older products. Moreover, a wide spread in the age classes used for blending reduces decay risk exposure.

Abstract

Abstract
Background and Objectives: Early detection of cognitive impairment is crucial for timely clinical interventions aimed at delaying progression to dementia. However, existing screening tools are not ideal for wide population screening. This study explores the potential of combining machine learning, specifically, one-class classification, with simpler and quicker motor-cognitive tasks to improve the early detection of cognitive impairment. Methods: We gathered data on gait, fingertapping, cognitive, and dual tasks from older adults with mild cognitive impairment and healthy controls. Using one-class classification, we modeled the behavior of the majority group (healthy controls), identifying deviations from this behavior as abnormal. To account for confounding effects, we integrated confound regression into the classification pipeline. We evaluated the performance of individual tasks, as well as the combination of features (early fusion) and models (late fusion). Additionally, we compared the results with those from two-class classification and a standard cognitive screening test. Results: We analyzed data from 37 healthy controls and 16 individuals with mild cognitive impairment. Results revealed that one-class classification had higher predictive accuracy for mild cognitive impairment, whereas two-class classification performed better in identifying healthy controls. Gait features yielded the best results for one-class classification. Combining individual models led to better performance than combining features from the different tasks. Notably, the one-class majority voting approach exhibited a sensitivity of 87.5% and a specificity of 75.7%, suggesting it may serve as a potential alternative to the standard cognitive screening test. In contrast, the two-class majority voting failed to improve the low sensitivities achieved by the individual models due to the underrepresentation of the impaired group. Conclusion: Our preliminary results support the use of one-class classification with confound control to detect abnormal patterns of gait, fingertapping, cognitive, and dual tasks, to improve the early detection of cognitive impairment. Further research is necessary to substantiate the method's effectiveness in broader clinical settings.

Abstract
In today's highly competitive fashion retail market, it is crucial to have accurate demand forecasting systems, namely for new products. Many experts have used machine learning techniques to forecast product sales. However, sales that do not happen due to lack of product availability are often ignored, resulting in censored demand and service levels that are lower than expected. Motivated by the relevance of this issue, we developed a two-stage approach to forecast the demand for new products in the fashion retail industry. In the first stage, we compared four methods of transforming historical sales into historical demand for products already commercialized. Three methods used sales-weighted averages to estimate demand on the days with stock-outs, while the fourth method employed an Expectation-Maximization (EM) algorithm to account for potential substitute products affected by stock-outs of preferred products. We then evaluated the performance of these methods and selected the most accurate one for calculating the primary demand for these historical products. In the second stage, we predicted the demand for the products of the following collection using Random Forest, Deep Neural Networks, and Support Vector Regression algorithms. In addition, we applied a model that consisted of weighting the demands previously calculated for the products of past collections that were most similar to the new products. We validated the proposed methodology using a European fashion retailer case study. The results revealed that the method using the Expectation-Maximization algorithm had the highest potential, followed by the Random Forest algorithm. We believe that this approach will lead to more assertive and better-aligned decisions in production management.

Abstract