AMIR 2019 — The 1st Interdisciplinary Workshop on Algorithm Selection and Meta-Learning in Information Retrieval — is over, and it was a full success. We had many interesting presentations and around 20 attendees. The following photos give some impression of the presentations:
We finalized the list of accepted papers and the schedule for the 1st Interdisciplinary Workshop on Algorithm Selection and Meta-Learning in Information Retrieval (AMIR) on 14 April 2019, Cologne, Germany. Out of 10 submissions, we accepted 3 full-papers and 2 short-papers/demos. We will also have two hands-on sessions about ASLib, auto-sklearn and Auto-PyTorch, and we are delighted to announce Dr. Marius Lindauer (University of Freiburg) as a keynote speaker (more details to follow soon). Here is the preliminary programme subject to minor changes. The latest programme can always be found at http://amir-workshop.org/amir-2019/for-attendees/programme-accepted-papers/. 08:00 Registration 09:00 Welcome Joeran Beel (Trinity College Dublin) and Lars Kotthoff (University of Wyoming) 09:10 Keynote Marius Lindauer (University of Freiburg) Title and Abstract: TBA Biography: Marius research focus lies on the performance tuning of any kind of algorithm (e.g., SAT solvers or machine learning algorithms) using cutting edge techniques from machine learning and optimization. A well-known, but also a tedious, time-consuming and error-prone way to optimize performance (e.g., runtime or prediction loss) is to tune the algorithm’s (hyper-) parameters in a manual way. To lift the burden on developers and users, Marius develops methods to automate the process of parameter tuning and algorithm selection for a given problem at hand (e.g., a machine learning dataset, or a set of SAT formulas). To this end, Marius provides ready-to-use, push-button software that enables users to optimize their software in an easy and efficient way. 10:10 Algorithm selection with librec-auto Masoud Mansoury and Robin Burke Due to the complexity of recommendation algorithms, experimentation on recommender systems has become a challenging task. Current recommendation algorithms, while powerful, involve large numbers of hyperparameters. Tuning hyperparameters for finding the best recommendation outcome often requires execution of large numbers of algorithmic experiments particularly when multiples evaluation metrics are considered. Existing recommender systems Read more…
The late-breaking-result submission page (EasyChair) opened today. We invite submissions relating to algorithm selection and meta-learning particularly in the field of information retrieval. Accepted formats include Full Papers (max. 12 pages + references), Short Papers / Posters (max. 6 pages + references), and Nectar papers (1 page + references). More details in our Call for Papers. Deadline is March 15, 2019.
We are delighted to announce Dr. Marius Lindauer as keynote speaker. Marius Lindauer is a junior research group lead at the University of Freiburg (Germany). His goal is to make the technology behind state-of-the-art research on artificial intelligence (AI) available to everyone. To this end, his research and tools aim at automating the development process of new AI systems. He received his M.Sc. and Ph.D. in computer science at the University of Potsdam (Germany), where he worked in the Potassco group. In 2014, he moved to Freiburg i.Br. (Germany) as a postdoctoral research fellow in the AutoML.org group. In 2013, he was one of the co-founders of the international research network COSEAL (COnfiguration and SElection of ALgorithms) and is nowadays a member of its advisory board. Besides organizing the first open algorithm selection challenge and winning several international AI competitions, he was a member of the team that won the first and second edition of the international challenge on automated machine learning. The title of Marius’ talk is “Automated Algorithm Selection: Predict which algorithm to use!” Abstract: To achieve state-of-the-art performance, it is often crucial to select a suitable algorithm for a given problem instance. For example, what is the best search algorithm for a given instance of a search problem; or what is the best machine learning algorithm for a given dataset? By trying out many different algorithms on many problem instances, developers learn an intuitive mapping from some characteristics of a given problem instance (e.g., the number of features of a dataset) to a well-performing algorithm (e.g., random forest). The goal of automated algorithm selection is to learn from data, how to automatically select a well-performing algorithm given such characteristics. In this talk, I will give an overview of the key ideas behind algorithm selection and different approaches Read more…
We have received 7 “standard” submissions and reviews are almost all in. In the next couple of days, we will discuss the reviews and make the final decisions. You can expect the notifications to be sent between the 4th and 6th of March. Keep in mind that we still accept late-breaking-result submissions until the 15th of March.
We finalized the call for papers, important dates, and submission instructions for the 1st Interdisciplinary Workshop on Algorithm Selection and Meta-Learning in Information Retrieval (AMIR).
We are delighted to announce the launch of our new website for AMIR — The 1st Interdisciplinary Workshop on Algorithm Selection and Meta-Learning in Information Retrieval. AMIR has been accepted as a workshop at the 41st European Conference on Information Retrieval (ECIR) in Cologne, Germany. Our goal is to bring together researchers from the fields of algorithm selection and meta-learning as well as information retrieval. We aim to raise the awareness of the algorithm selection problem in the IR community; identify the potential for automatic algorithm selection in information retrieval; and explore possible solutions for this context. In particular, we will explore to what extent existing solutions to the algorithm selection problem from other domains can be applied in information retrieval, and also how techniques from IR can be used for automated algorithm selection and meta-learning. AMIR is held on 14 April 2019 in Cologne, Germany. Follow us on Twitter to receive regular updates.