HEC Montréal

About: HEC Montréal is a education organization based out in Montreal, Quebec, Canada. It is known for research contribution in the topics: Context (language use) & Vehicle routing problem. 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.

Language GANs Falling Short

Abstract: Traditional natural language generation (NLG) models are trained using maximum likelihood estimation (MLE) which differs from the sample generation inference procedure. During training the ground truth tokens are passed to the model, however, during inference, the model instead reads its previously generated samples - a phenomenon coined exposure bias. Exposure bias was hypothesized to be a root cause of poor sample quality and thus many generative adversarial networks (GANs) were proposed as a remedy since they have identical training and inference. However, many of the ensuing GAN variants validated sample quality improvements but ignored loss of sample diversity. This work reiterates the fallacy of quality-only metrics and clearly demonstrate that the well-established technique of reducing softmax temperature can outperform GANs on a quality-only metric. Further, we establish a definitive quality-diversity evaluation procedure using temperature tuning over local and global sample metrics. Under this, we find that MLE models consistently outperform the proposed GAN variants over the whole quality-diversity space. Specifically, we find that 1) exposure bias appears to be less of an issue than the complications arising from non-differentiable, sequential GAN training; 2) MLE trained models provide a better quality/diversity trade-off compared to their GAN counterparts, all while being easier to train, easier to cross-validate, and less computationally expensive.

Models and a Branch-and-Cut Algorithm for Pickup and Delivery Problems with Time Windows

Abstract: In the pickup and delivery problem with time windows (PDPTW), capacitated vehicles must be routed to satisfy a set of transportation requests between given origins and destinations. In addition to capacity and time window constraints, vehicle routes must also satisfy pairing and precedence constraints on pickups and deliveries. This paper introduces two new formulations for the PDPTW and the closely related dial-a-ride problem (DARP) in which a limit is imposed on the elapsed time between the pickup and the delivery of a request. Several families of valid inequalities are introduced to strengthen these two formulations. These inequalities are used within a branch-and-cut algorithm which has been tested on several sets of instances for both the PDPTW and the DARP. Instances with up to eight vehicles and 96 requests (192 nodes) have been solved to optimality.

The Influence of a Structurationist View on Strategic Management Research

Abstract: In this article, strategic management research using structuration theory from 1995 to 2000 is reviewed. I describe and analyse the theoretical articulations adopted to make sense of strategy using a structurationist view. I found that, instead of being applied as the sole theoretical foundation, Giddens' propositions have been incorporated into other perspectives, the effects of which should be known by researchers looking for theoretical frameworks that avoid dichotomist thinking. The paper draws on the effects that structurationist arguments may produce regarding classical oppositions such as micro/macro and voluntarist/determinist. Its main contribution is to show how theoretical complementarities using structuration theory are promising avenues of research in the strategic management field. It also suggests that, although other alternatives of avoiding dichotomist logic exist, making a choice among them is more a question of ontological affinity than of making the 'better choice' among competing accounts. There are several routes to advance the understanding of the possibilities of human choice.