Abstract
Machine Learning systems are generally thought of as fully automatic. However, in recent years, interactive systems in which Human experts actively contribute towards the learning process have shown improved performance when compared to fully automated ones. This may be so in scenarios of Big Data, scenarios in which the input is a data stream, or when there is concept drift. In this paper we present a system for supporting auditors in the task of financial fraud detection. The system is interactive in the sense that the auditors can provide feedback regarding the instances of the data they use, or even suggest new variables. This feedback is incorporated into newly trained Machine Learning models which improve over time. In this paper we show that the order by which instances are evaluated by the auditors, and their feedback incorporated, influences the evolution of the performance of the system over time. The goal of this paper is to study of different instance selection strategies for Human evaluation and feedback can improve the learning speed. This information can then be used by the system to determine, at each moment, which instances would improve the system the most, so that these can be suggested to the users for validation.

Abstract
Sustainability and eco-efficiency have been researched in multiple scientific papers since the last years. However the literature is not so abundant when applying those concepts to industrial assembly processes. This paper presents an innovate methodology to optimize aerospace assembly processes. Authors propose the introduction of a new element, the eco-efficiency, along with the traditional criteria, cost and time, currently used for optimization. Using a large Aero-Structure as an industrial case of study, the methodology analyzes the eco-efficiency of an assembly process in connection with a Life Cycle Assessment (LCA) to compute the environmental impact. Results are shown in a dashboard along with the relevant Key Process Indicator (KPI) to help the engineers to select the best assembly process. © 2020, IFIP International Federation for Information Processing.

Abstract

Abstract

Abstract
Facility Location embodies a class of problems concerned with locating a set of facilities to serve a geographically distributed population of customers at minimum cost. We address the classical Capacitated Facility Location Problem (CFLP) in which the assignment of facilities to customers must ensure enough facility capacity and all the customers must be served. This is a well-known NP-hard problem in combinatorial optimization that has been extensively studied in the literature. Due to the difficulty of the problem, significant research efforts have been devoted to developing advanced heuristic methods aimed at finding high-quality solutions in reasonable computational times. We propose a Relaxation AdaptiveMemory Programming (RAMP) approach for the CFLP. Our method combines lagrangean subgradient search with an improvement method to explore primal-dual relationships to create advanced memory structures that integrate information from both primal and dual solution spaces. The algorithm was tested on the standard ORLIB dataset and on other very large-scale instances for the CFLP. Our approach efficiently found the optimal solution for all ORLIB instances and very competitive results for the large-scale ones. Comparisons with current best-performing algorithms for the CFLP show that our RAMP algorithm exhibits excellent results.

Abstract
We propose a Relaxation Adaptive Memory Programming (RAMP) algorithm for the solution of the Single Source Capacitated Facility Location Problem (SSCFLP). This problem considers a set of possible locations for opening facilities and a set of clients whose demand must be satisfied. The objective is to minimize the cost of assigning the clients to the facilities, ensuring that all clients are served by only one facility without exceeding the capacity of the facilities. The RAMP framework efficiently explores the relation between the primal and the dual sides of combinatorial optimization problems. In our approach, the dual problem, obtained through a lagrangean relaxation, is solved by subgradient optimization. Computational experiments of the effectiveness of this approach are presented and discussed.