Title: Reverse AD for an Array Language with Nested Parallelism - Redundant Execution: A Practical Backpropagator

Speaker affiliation: Prof. Dr. Cosmin Oancea, University of Copenhagen, Denmark.

Talk abstract: Automatic differentiation (AD) is a practical way for computing derivatives of functions that are expressed as programs. AD has been recognized as one of the key pillars of the current machine learning (ML) revolution, and has key applications in other domains such as financial algorithms, computational fluid dynamics, atmospheric sciences, and engineering design optimization. This talk presents a technique for applying reverse AD on a higher-order functional array language, which allows programs to be written as a nested composition of parallel operators such as map, reduce, scan (also together with loops with in-place updates).

Short bio: Cosmin Oancea is an Associate Professor with the Department of Computer Science (DIKU), University of Copenhagen. His main research interests lie in the general field of programming language design and implementation, with focus on compiler optimizations aimed at parallel execution on highly-parallel hardware.

Talk title: Developing and implementing contextualised Sales and Operations Planning (S&OP) designs - experiences from a PROBRAL project.

Speaker affiliation: Tobias Kreuter, Ph.D. candidate at the Chair for Information Systems & Supply Chain Management, University of Münster.

Talk abstract: Tobias speaks about his recent research in relation to the PROBRAL project “Theoretical understanding and practical implementation guidance for Sales and Operations Planning (S&OP)”, which is conducted in collaboration with the Pontifical Catholic University of Rio de Janeiro and the Federal University of the State of Rio de Janeiro. In particular, the talk focuses on the development and implementation of contextualised S&OP designs. Tobias presents a novel approach for S&OP contextualisation, rooted in Enterprise Architecture Management and contingency theory.

Short Bio: Tobias is a researcher and Ph.D. candidate at the Chair for Information Systems & Supply Chain Management at the European Research Center for Information Systems (ERCIS), University of Münster. He is part of the above-mentioned PROBRAL project (http://go.wwu.de/ptp-h).

Talk title: Ethics in AI: A Challenging Task

Speaker affiliation: Ricardo Baeza-Yates, Institute for Experiential AI, Northeastern University.

Talk abstract: In the first part we cover five current specific challenges through examples: (1) discrimination (e.g., facial recognition, justice, sharing economy, language models); (2) phrenology (e.g., biometric based predictions); (3) unfair digital commerce (e.g., exposure and popularity bias); (4) stupid models (e.g., Signal, minimal adversarial AI) and (5) indiscriminated use of computing resources (e.g., large language models). These examples do have a personal bias but set the context for the second part where we address four generic challenges: (1) too many principles (e.g., principles vs. techniques), (2) cultural differences (e.g., Christian vs. Muslim); (3) regulation (e.g., privacy, antitrust) and (4) our cognitive biases. We finish discussing what we can do to address some of these challenges in the near future.

Short bio: Ricardo Baeza-Yates is Research Professor at the Institute for Experiential AI of Northeastern University. He is also part-time professor at Universitat Pompeu Fabra in Barcelona and Universidad de Chile in Santiago. Before, he was VP of Research at Yahoo Labs, based in Barcelona, Spain, and later in Sunnyvale, California, from 2006 to 2016.

Speaker: Dr. Doina Bucur, University of Twente, The Netherlands.

Abstract: Information (of influence) diffuses via links in a social network, and, even assuming that the network structure is relatively static, the size of an information cascade is hard to estimate well, both over a model and over a real social network. We cover models of information diffusion, then look at methods to predict ‘important’ nodes in the network, for example single nodes which would be able to influence many others single-handedly (in other words, maximise influence), or small sets of nodes which would do so in combination. The definition of node ‘importance’ changes with the case study, and may also require the node to minimise the size of the diffusion (for example in epidemics).

Speaker: Dr. Sirkka L. Jarvenpaa, The University of Texas at Austin, USA.

Abstract: Global virtual teams represent temporary work systems that are assembled for a joint task, performed by team members who collaborate primarily via digital technologies. Team members span geography and culture and often have only a narrow period of shared work hours. Within highly constrained temporal spaces, team members coordinate and collaborate on joint tasks with many task interdependencies, requiring constant back-and-forth workflows among members. Leveraging various synchronous and asynchronous virtual communication modes, the teams must communicate effectively to prevent prolonged misunderstandings and work delays. The virtual team environment is challenging because virtual space reduces opportunities for team members to grasp important aspects of the actual social surroundings of the members that are critical for understanding.

Speaker: Prof. Dr. Margunn Aanestad, Department of Information Systems, University of Agder, Kristiansand, Norway.

Abstract: How do configurations of humans and algorithms evolve as firms adopt artificial intelligence (AI) capabilities, and what are the implications for work and organization? We explored these questions through a two-year long case study of an organization in the international maritime trade that introduced automated algorithmic support for data analysis and prediction work.