An ensemble method for concept drift in nonstationary environment

Most statistical and data-mining algorithms assume that data come from a stationary distribution. However, in many real-world classification tasks, data arrive over time and the target concept to be learned from the data stream may change accordingly. Many algorithms have been proposed for learning drifting concepts. To deal with the problem of learning when the distribution generating the data changes over time, dynamic weighted majority was proposed as an ensemble method for concept drift. Unfortunately, this technique considers neither the age of the classifiers in the ensemble nor their past correct classification. In this paper, we propose a method that takes into account expert's age as well as its contribution to the global algorithm's accuracy. We evaluate the effectiveness of our proposed method by using m classifiers and training a collection of n-fold partitioning of the data. Experimental results on a benchmark data set show that our method outperforms existing ones.     

Decentralised bilevel model for shared inventory management

In this study, we consider the case of a large corporation that owns a warehouse and two concurrent chains. The company opted for shared inventory policy. The warehouse is looking into minimising its inventory cost and at the same time reducing the fluctuation of the personnel. The chains, independently, are trying to minimise their inventory costs. The warehouse plays the role of a leader in optimising its objectives and, as followers, the two chains try to satisfy their objectives. The problem is formulated into a decentralised nonlinear bilevel programming problem. We also consider the fuzziness of some parameters due to the imprecise available information. We propose the mathematical model and we solve it for our specific case.     

An Attempt in Modeling the Labor Supply Process of Married Women in Tunisia

ABSTRACT

This article is aimed at determining the appropriate model for a married woman’s labor supply process in the presence of strategic interaction. We here compare two approaches: a traditional model and a strategic model. We try to estimate the process using Probit and Strategic models and specify, via structural equations, the advantage of the strategic model in the study of strategic data. Using the two models on individual Tunisian data enables us to note that as contrary to the woman expectations in a Probit model, the financial side is not a determining factor in the husband’s reaction once his wife gave up her work. This result has been observed only via the strategic model because it takes into account the presence of effect of strategic interaction between a woman’s decision and her husband’s reaction.