• Employee Views of Leveraged Buy-Out Transactions

    Date: vendredi 4 novembre 2022 Heure: Hybride, Séminaire organisé conjointement avec Quantact et le département de finance Résumé: This paper offers a comprehensive view of employee satisfaction: we make use of 700,000 ratings, in addition to 500,000 written reviews posted by employees of all ranks, in different industries, types of companies, and in companies that underwent different types of ownership changes. Employee satisfaction is lower following a company acquisition; more so when it is a Leveraged Buy-Out (LBO).
  • The Leverage Effect and Propagation

    Date: vendredi 10 juin 2022 Heure: 11h00-12h00 Hybride, Salle Sony et Zoom Séminaire organisé conjointement avec Quantact Résumé: This paper proposes a new way to measure the leverage effect and its propagation over time. We also show that, with respect to the newly proposed measure, common volatility models like the GJRGARCH, the Exponential GARCH, and the asymmetric SV can be inaccurate to correctly represent the leverage effect and its propagation for financial time series.
  • Efficient Estimation of Bid-Ask Spreads from Open, High, Low, and Close Prices

    Date: mardi 10 mai 2022 Heure: 12h00-13h00 Salle Transat Séminaire brownbag organisé conjointement avec le département de finance. Résumé: This paper formally derives an efficient estimator of the bid-ask spread from open, high, low, and close prices. The estimator is asymptotically unbiased and optimally combines the full set of price data to minimize the estimation variance. In absence of quote data, it delivers the most accurate estimates of bid-ask spreads theoretically, numerically, and empirically.
  • Learning and reasoning with constraint solving

    Date: mardi 10 mai 2022 Heure: 11h00-12h00 En ligne (Zoom) Webinaire organisé conjointement avec le GÉRAD Résumé: Industry and society are increasingly automating processes, which requires solving constrained optimisation problems. This includes vehicle routing, demand-response planning, rostering and more. To find not just optimal solutions, but also ‘desirable’ solution by the end user, it is increasingly important to offer tools that automatically learn from the user and the environment and that support the constraint modelling in interpretable ways.
  • Toward social welfare and fairness in kidney exchange programs

    Date: jeudi 28 avril 2022 Heure: 11h00-12h00 Salle Manuvie, bâtiment Côte-Sainte-Catherine Séminaire webdiffusé, organisé conjointement avec la Chaire de théorie des jeux et gestion Résumé: Matching markets are part of our daily lives, appearing on online platforms, school admissions and health systems. Their study attracts the interest of optimizers and game theorists. In this talk, we will focus on a particular matching market, the kidney exchange program (KEP), where combinatorial optimization and game theory play an important role.
  • Comparative Probability Metrics: Using Posterior Probabilities to Account for Practical Equivalence in A/B tests

    Date: mercredi 23 mars 2022 Heure: 15h00-16h00 Résumé: Online controlled experiments (i.e., A/B tests) have become an extremely valuable tool used by internet and technology companies for purposes of advertising, product development, product improvement, customer acquisition, and customer retention to name a few. The data-driven decisions that result from these experiments are typically informed by null hypothesis significance tests and analyses based on p-values. However, attention has recently been drawn to the shortcomings of hypothesis testing, and an emphasis has been placed on the development of new methodologies that overcome these shortcomings.
  • Differentially private inference via noisy optimization

    Date: vendredi 25 février 2022 Heure: 15h00-16h00 Webinaire organisé conjointement avec le département de mathématiques et de statistique de l’Université Laval Résumé: We propose a general optimization-based framework for computing differentially private M-estimators and a new method for constructing differentially private confidence regions. Firstly, we show that robust statistics can be used in conjunction with noisy gradient descent or noisy Newton methods in order to obtain optimal private estimators with global linear or quadratic convergence, respectively.
  • Size Distribution of Firms and Strategic Investments in Large Markets: A Stochastic Mean Field Game Approach

    Date: vendredi 5 novembre 2021 Heure: 14h00-15h00 Résumé: This paper analyzes how firm size heterogeneity distribution affects capacity investments and profits in large competitive markets with both idiosyncratic and aggregate uncertainty shocks. We use a mean field game approach where firms’ sizes are heterogeneously distributed and decisions on capacity investments (addition) or disinvestments (withdrawal) are optimally made over time to enhance prospects of earning profits under uncertainty. A mean field game framework allows for consideration of markets with a large population of interacting firms.
  • Opinionated practices for teaching reproducibility: motivation, guided instruction and practice

    Date: vendredi 29 octobre 2021 Heure: 15h30-16h30 Résumé: In the data science courses at the University of British Columbia, we define data science as the study, development and practice of reproducible and auditable processes to obtain insight from data. While reproducibility is core to our definition, most data science learners enter the field with other aspects of data science in mind, for example predictive modelling, which is often one of the most interesting topic to novices.
  • Selective inference on trees

    Date: lundi 4 octobre 2021 Heure: 15h-16h Résumé: The ever-increasing scope and scale of data collection has shifted the focus of data collection away from testing pre-specified hypotheses and towards hypothesis generation. Researchers are often interested in performing exploratory data analysis on a data set to generate hypotheses, then to validate those hypotheses in that same data via tests of significance. Unfortunately, this type of “double-dipping” can lead to extremely inflated type I error rates.