Hospital service levels during drug shortages: Stocking and transshipment policies for pharmaceutical inventory

Journal of Global Optimization - In this study, we consider a health network that faces uncertain supply disruptions in the form of regional, nationwide, or worldwide drug shortages. Each hospital...     

Raman spectroscopy and support vector machines for quick toxicological evaluation of titania nanoparticles

With the rapid development of nanotechnology products, there is a significant concern on the adverse effects that might be associated with them. Traditional biological assays are typically used to asses the toxicity in vitro. There are, however, questions regarding the suitability of these assays for this purpose, mainly due to the potential interaction of the particles with the utilized dyes. In addition, this process can be costly and time consuming, as a large number of different assays have to be used. To address some of these issues, Raman spectroscopy is used in this study to investigate the particle‐cell interactions. The spectrum of a living cell is a very complex and rich collection of data directly related to its chemical composition. To enhance the data resolution and make the detection of toxicity more robust, data‐mining techniques have been deployed. Furthermore, data‐mining techniques enable full automation of the entire process, minimizing user input. The Raman spectroscopy successfully evaluated the toxicity of TiO₂ nanoparticles by both the peak‐by‐peak analysis and with the implementation of support vector machines. The particles were found to display cytotoxicity after 36 h of exposure. The results were confirmed by MTT (3‐(4,5‐Dimethylthiazol‐2‐Yl)‐2,5‐Diphenyltetrazolium Bromide) assay and are in agreement with the existing literature on the subject. Overall, Raman spectroscopy appears to be among the very few techniques that exhibit low levels of interferences (obscuration, fluorescence, emission, etc.) from the particle addition. Since it does not rely on biomarkers, it can be used in situ for an extended period with minimal effects on the cellular biochemistry. Copyright © 2011 John Wiley & Sons, Ltd.     

Robust support vector machines for multiple instance learning

This paper presents the multiple instance classification problem that can be used for drug and molecular activity prediction, text categorization, image annotation, and object recognition. In order to model a more robust representation of outliers, hard margin loss formulations that minimize the number of misclassified instances are proposed. Although the problem is $\mathcal{NP}$ -hard, computational studies show that medium sized problems can be solved to optimality in reasonable time using integer programming and constraint programming formulations. A three-phase heuristic algorithm is proposed for larger problems. Furthermore, different loss functions such as hinge loss, ramp loss, and hard margin loss are empirically compared in the context of multiple instance classification. The proposed heuristic and robust support vector machines with hard margin loss demonstrate superior generalization performance compared to other approaches for multiple instance learning.     

An EOQ model with deteriorating items and self-selection constraints

Optimization Letters - In this paper, we consider a store that sells two vertically differentiated items that might substitute each other. These items do not only differ in quality and price, but...     

Integrated market selection and production planning: complexity and solution approaches

Emphasis on effective demand management is becoming increasingly recognized as an important factor in operations performance. Operations models that account for supply costs and constraints as well as a supplier’s ability to influence demand characteristics can lead to an improved match between supply and demand. This paper presents a class of optimization models that allow a supplier to select, from a set of potential markets, those markets that provide maximum profit when production/procurement economies of scale exist in the supply process. The resulting optimization problem we study possesses an interesting structure and we show that although the general problem is ${\mathcal{NP}}$ -complete, a number of relevant and practical special cases can be solved in polynomial time. We also provide a computationally very efficient and intuitively attractive heuristic solution procedure that performs extremely well on a large number of test instances.