HEC Montréal

About: HEC Montréal is a education organization based out in Montreal, Quebec, Canada. It is known for research contribution in the topics: Vehicle routing problem & Corporate governance. The organization has 1221 authors who have published 5708 publications receiving 196862 citations. The organization is also known as: Ecole des Hautes Etudes Commerciales de Montreal & HEC Montreal.

Models and algorithms for the dynamic-demand joint replenishment problem

Abstract: This paper considers the problem of coordinated ordering of items having deterministic but time-varying demands where there is a common ordering cost if one or more of these items are ordered in addition to individual items ordering costs. Two new integer linear-programming formulations are presented and compared with the classical formulation. Several well-known heuristics are described and a new improvement procedure is proposed. The relative performance of these heuristics is assessed. Results show the superiority of the new formulations and of the new improvement procedure.

Incentive Mechanisms for Safe Driving: A Comparative Analysis with Dynamic Data

Abstract: Road safety policies often use incentive mechanisms based on traffic violations to promote safe driving. Examples of mechanisms are fines, experience rating and point-record driving licenses. We analyse the effectiveness of these mechanisms in promoting safe driving. We derive their theoretical properties with respect to contract time and accumulated demerit points. These properties are tested empirically with data from the Quebec public insurance plan. We find evidence of moral hazard, which means that drivers who accumulate demerit points become more careful because they are at threat of losing their license. The insurance rating scheme introduced in 1992 reduced the frequency of traffic violations by 15%. We use this result to derive monetary equivalents for traffic violations and license suspensions.

Online Fast Adaptation and Knowledge Accumulation (OSAKA): a New Approach to Continual Learning

Abstract: Continual learning agents experience a stream of (related) tasks. The main challenge is that the agent must not forget previous tasks and also adapt to novel tasks in the stream. We are interested in the intersection of two recent continual-learning scenarios. In meta-continual learning, the model is pre-trained using meta-learning to minimize catastrophic forgetting of previous tasks. In continual-meta learning, the aim is to train agents for faster remembering of previous tasks through adaptation. In their original formulations, both methods have limitations. We stand on their shoulders to propose a more general scenario, OSAKA, where an agent must quickly solve new (out-of-distribution) tasks, while also requiring fast remembering. We show that current continual learning, meta-learning, meta-continual learning, and continual-meta learning techniques fail in this new scenario. We propose Continual-MAML, an online extension of the popular MAML algorithm as a strong baseline for this scenario. We show in an empirical study that ContinualMAML is better suited to the new scenario than the aforementioned methodologies including standard continual learning and meta-learning approaches.