Just two years after receiving the best paper award at the 15th International Conference on Parallel Problem Solving from Nature (PPSN 2018), PhD student and Vanier Scholar Yasha Pushak and Professor Holger Hoos received another best paper award at the 22nd international Genetic and Evolutionary Computation Conference (GECCO 2020) for a line of research on automated algorithm configuration. Their second paper, titled “Golden Parameter Search: Exploiting Structure to Quickly Configure Parameters in Parallel” was the single winner of the ECOM Track at GECCO 2020, determined by voting of the audience.
Important and challenging computation problems are typically solved through the use of highly parameterized algorithms. Finding high-quality parameter settings for the algorithm can often improve their performance by several orders of magnitude. In the last two decades, automated algorithm configuration has emerged as a hot topic of research.
State of the art algorithm configurators rely on powerful search heuristics to explore the space of parameter configurations; however, they are typically quite costly, often requiring days or even weeks to find high-quality parameter configurations. These methods assume that finding high-quality parameter configurations is challenging. Consider, for example, the problem of trying to climb to the highest point of a rugged mountain with a large number of peaks in a dense fog: If you always climb up the steepest hill, you will likely get stuck at a peak that is far from the highest on the mountain. To avoid this, most configurators spend time exploring the entire “landscapes” of the algorithm configuration problem — including regions that don’t initially appear promising, just to be certain they haven’t missed anything.
In their 2018 PPSN award-winning paper, “Algorithm Configuration Landscapes: More Benign than Expected?” Yasha Pushak and Holger Hoos hypothesized that the landscapes induced by algorithm configuration problems are likely much simpler and more structured than previously assumed. By way of analogy, consider the task of choosing a temperature at which to bake a cake: If the temperature is too high the cake will burn, but if it is too low the cake won’t bake. Therefore, finding the optimal temperature simply corresponds to trading off between “too high” and “too low”. Yasha Pushak and Holger Hoos developed a statistical procedure to test for similar structure in the response of algorithm parameters. They applied their procedure to a large set of parameters and were unable to reject their hypotheses in 99.5% of the cases they studied. Read more about algorithm configuration landscapes here.
Capitalizing on these insights, Yasha Pushak and Holger Hoos developed a new automated algorithm configuration procedure: Golden Parameter Search (GPS). GPS is the first of a new generation of automated algorithm configuration procedures designed to exploit structural properties of algorithm configuration landscapes. In their 2020 GECCO award-winning paper, they showed that GPS was often capable of finding similar-or better-quality parameter configurations using a fraction of the computational budget required by several other long-standing, state-of-the-art algorithm configuration procedures. Read more about GPS here.
Traffic congestion tends to be bad for the environment, the economy, and above all: the drivers’ moods. As such, it is a worthwhile cause to pursue improvements for; in particular, being computer- and data scientists, using data-driven methods to try to alleviate this problem seemed a particularly exciting approach. This is just what we (Laurens Arp, Dyon van Vreumingen, Daniela Gahwens and Mitra Baratchi) published a paper for in the IEEE MDM 2020 proceedings. The project, which started as a course project for Mitra’s Urban Computing course at Leiden University, evolved first into a short paper submission to the Netmob Future Cities Challenge (FCC), and subsequently into a full paper submission to the MDM conference.
The main idea of the method proposed was to redistribute traffic by imposing external costs (or rewards) to specific road segments. We used a movement dataset for Tokyo, provided by Foursquare for the FCC, from which we could derive which proportion of drivers would want to go to and from specific parts of the city. By combining this with a macro-scale traffic flow model (Greenshields), we were able to compute the occupancy of specific roads and the total travel time this resulted in. This model could then be used as an objective function for black-box optimisation; we would optimise the cost parameters of road segments so that the optimal number of desired routes got redirected to alternative roads such that the overall traffic time was minimised.
The amount of improvement (under the Greenshields model) we were able to achieve was highly dependent on the number of cars we estimated would be on the road at the same time. The best improvement achieved was 13.15% for a little over 13 500 cars (925 hours), and the worst was 1.35% for just under 113 000 cars (437 hours). Interestingly, even the relatively modest improvements, occurring for large amounts of cars, could still be meaningful, because there are more drivers to benefit from them. We also added a few fairness analyses to the paper, the results of which did not seem to indicate any unfair disadvantage to individual drivers. You can find the pre-recorded conference presentation here.
We hope that our paper will be able to contribute to the perpetually on-going efforts worldwide of causing less damage to the environment, boosting the economy, and perhaps also helping some drivers’ moods.
How could AI techniques be used to best improve the living standards of people around the world? This is the main question of interest for Laurens Arp, a Master student who joined the ADA research group in November 2019. He is currently working on his Master Thesis under the supervision of Mitra Baratchi and Holger Hoos. The project is about the data-driven evaluation and optimization methods of geographical regions.
The main focus of the project will be on (spatial) representation learning. Current methods would not sufficiently address the problem yet, as most approaches will either focus too much on spatial structure instead of how this structure affects the features of a neighborhood, or are aimed too much at encoding similarity rather than interaction. Once a suitable representation has been found, machine learning and deep learning could be used to automatically learn the relationship between geographical features and the measures one might like to use to evaluate a region. If the resulting model is sufficiently accurate, it could then be used to rate the quality of region configurations generated by an optimization algorithm, allowing for the optimization of the development of the region.
Zhou joined the ADA Research Group in September 2019 as a visiting PhD student for a period of one year. He has started his PhD in March 2017 under the supervision of dr. Gangquan Si at School of Electrical Engineering, Xi’an Jiaotong University.
In his research he focuses on time series data mining techniques, including pre-processing, representation, classification, and prediction. His work has been successfully applied to a project titled “Research of Data Mining Technique and Development of Intelligent Data Management System for Electrical Equipment”, supported by China Southern Power Grid. In this project, he tries to make full use of massive data collected from various tests and online monitoring systems, aimed at providing accurate evaluation and prediction of electrical equipment status.
Zhou is currently working on online time series segmentation with the purpose of dimension reduction, especially on how to apply Automated Machine Learning methods for this task. During his visit, he will be closely supervised by dr. Mitra Baratchi and prof. Holger Hoos.
Daniël Fokkinga joined the ADA research group as a Master’s student in September of 2018. His research is jointly supervised by Anna Louise Latour and Marie Anastacio and applies Automated Algorithm Configuration (AAC) on a pipeline for solving Stochastic Constraint Optimisation Problems (SCOPs).
An example of such a SCOP is the viral marketing problem. We are given a social network where an edge between two nodes represents a probabilistic communication relationship between two people in the network. We want to launch a new product by handing out a limited number of free samples to people in the network. We hope that they will like our product and spread the word to turn their acquaintances into buyers of our product. To whom in the network do we give the free samples, in order to maximise the expected number of people that will buy our product after they’ve heard about it from others? Who are the most influential people in our network?
Recently, a new pipeline for solving SCOPs was proposed. While it has shown its merit in a proof of principle, the different steps in the pipeline have not been optimised yet. While alternative design choices for elements in the pipeline have been proposed, their influence on the performance of the pipeline over a wide range of problems has not been extensively explored.
Daniël’s focus is on exposing these alternatives as parameters to optimise the pipeline for different applications, using Automated Algorithm Configuration.
Automated Algorithm Configuration allows a user to leave the tuning of parameters and thus the decision of the best approach to follow for parts of the full pipeline to a computer. Given a set of example instances, a configurator finds the combination of parameter values that is most likely to perform well on a new instance.
With this research Daniël hopes to show another successful application of Algorithm Configuration in a new field and improve the performance of previously developed algorithms.
Ada research group welcomes another guest researcher, Yi Chu. She will stay with Ada research group for a year.
Yi is a PhD candidate at the Institute of Computing Technology, Chinese Academy of Sciences. She received her master’s degree in computer technology from Beijing University on Posts and Telecommunications in 2014.
Yi’s research interests include heuristic algorithms for NP-hard problems and automatic algorithm design using optimization and learning techniques. Currently, she is working on using programming by optimization to improve the performance of heuristic algorithms for solving the maximum clique problem.
ADA research group welcomes Yanyan Xu, a new visiting scholar!
Yanyan Xu is a visiting scholar at LIACS. She joined ADA Research Group in September 2018. Yanyan holds M.Sc. and B.Sc. degrees from Sun Yat-sen University. She also received her Ph.D. degree from the Institute of Software, Chinese Academy of Sciences. Since then, she has been working at the School of Information Science and Technology of Beijing Forestry University. Currently, she holds an associate professor position there.
Yanyan Xu has a broad interest in the areas of artificial intelligence and algorithm design. She is particularly interested in pattern recognition and deep learning, as well as, heuristic algorithms in robotics and formal methods. Currently, she is working on combining formal methods with deep learning.