Last week I went to Budapest to present the paper “An Ant-Based Rule for UMDA’s Update Strategy” in the 10th European Conference on Artificial Life (ECAL 2009). ECAL is one of the leading congresses in the area and some of the most relevant work in the Artificial Life research field is presented there in first hand. It is held every two years and this time the capital of Hungary was chosen to host the event. The Academy of Sciences, in Roosevelt tér (square), on the banks of the Danube and with a perfect view on the Castle and the hills of Buda was ECAL’s headquarters for 4 days.
Only 30% of the accepted papers were selected for oral presentation. The remaining was scheduled for poster sessions (although all the accepted papers were published in full-length in two LNCS volumes) that lasted…the whole day! I cannot understand why not all the congresses follow a line similar to PPSN (a poster-only congress, with 90 minutes sessions) when it comes to poster sessions, but ECAL’s strategy is, my opinion, particularly ineffective and exhausting.
ECAL 2009, Budapest, Academy of Sciences
As for our paper, it presents a study on an alternative update strategy for the Univariate Marginal Distribution Algorithm based on the ACO computational paradigm and first presented here. The aim is to control the balance between exploration and exploitation in order to avoid diversity loss, reduce the optimal population size and improve the scalability of the algorithm on hard problems. The results confirmed the hypothesis. This is the abstract:
This paper investigates an update strategy for the Univariate Marginal Distribution Algorithm (UMDA) probabilistic model inspired by the equations of the Ant Colony Optimization (ACO) computational paradigm. By adapting ACO’s transition probability equations to the univariate probabilistic model, it is possible to control the balance between exploration and exploitation by tuning a single parameter. It is expected that a proper balance can improve the scalability of the algorithm on hard problems with bounded difficulties and experiments conducted on such problems with increasing difficulty and size confirmed these assumptions. These are important results because the performance is improved without increasing the complexity of the model, which is known to have a considerable computational effort.
Despite the intuition that the same population size is not needed throughout the run of an Evolutionary Algorithm (EA), most EAs use a fixed population size.
This paper presents an empirical study on the possible benefits of a Simple Variable Population Sizing (SVPS) scheme on the performance of Genetic Algorithms (GAs). It consists in decreasing the population for a GA run following a predetermined schedule, configured by a speed and a severity parameter. The method uses as initial population size an estimation of the minimum size needed to supply enough building blocks, using a fixed-size selectorecombinative GA converging within some confidence interval toward good solutions for a particular problem. Following this methodology, a scalability analysis is conducted on deceptive, quasi-deceptive, and non-deceptive trap functions in order to assess whether SVPS-GA improves performances compared to a fixed-size GA under different problem instances and difficulty levels. Results show several combinations of speed-severity where SVPS-GA preserves the solution quality while improving performances, by reducing the number of evaluations needed for success.
In this paper we analyse the robustness of a Peer-to-Peer (P2P) Evolutionary Algorithm (EA) subject to the following dynamics: peers leave the system independently from each other causing a collective effect known as churn. The algorithm has been designed to tackle large instances of computationally expensive problems and, in this paper, we will assess its behavior under churn. To that end, we have performed a scalability analysis in five different scenarios using the Massively Multimodal Deceptive Problem as a
benchmark. In all cases, the P2P EA reaches the success criterion without a penalty on the response time. The key to the algorithm robustness is to ensure enough peers at the beginning of the experiment. Some of them leave but those that remain are enough to guarantee a reliable convergence.
I returned from Brussels a couple of days ago, where I went to present KANTS: Artificial Ant System for Classification (the model was already described here) at the 6th International Conference on Ant Colony Optimization and Swarm Intelligence. ANTS 2008 is similar to PPSN, with most of the papers being presented at poster sessions (only a few are chosen for oral presentation). This procedure works well when the poster sessions are not just a minor event of the congress, thrown to a distant room in the hotel/university where nobody even bothers to go, or scheduled to the end of a long day. ANTS sessions were well organized and every poster had an assigned space. My presentation was scheduled to the last day of the congress, when most of people had already packed for their trip back home, but nonetheless the session went well, with lots of people wandering around the room, clearly interested in the works. KANTS got the attention of some audience, and I think they were quite impressed by the simplicity (and efficiency) of the idea. The inevitable question arouse (are you planning to test KANTS on a real-world problem?) and this time we can answer yes, we are not only planning to do it, but we are already working on it (later we will report on those experiments). In the same section, another swarm-clustering was presented. I saw the poster, and the results on clustering appear to be quite good (but the system does not perform classification). I haven’t read the paper (as a matter of fact, it is published as an extended abstract), but I was able to realize that the algorithm is little bit complex, simulating the behavior of three different species: ants, birds and spiders.
A week before I was in Barcelona, at the 8th International Conference on Hybrid Intelligent Systems (HIS 2008), presenting the paper Tracking Extrema in Dynamic Fitness Functions with Dissortative Mating Genetic Algorithms. It is quite a different work, and more related to my thesis’ subject, bio-inspired computation on dynamic environments. It describes the experiments performed with an adaptive dissortative mating GA (ADMGA) on Dynamic Optimization Problems. Dissortative mating appears frequently in nature and refers to the occurrence of mating between dissimilar individuals more often than expected by chance. It maintains genetic diversity at a higher level, thus increasing the exploration stage of the algorithm. Dynamic fitness functions are more sensible to genetic diversity than static ones, and so dissortative mating is a good candidate to deal with that kind of problems. The paper describes mainly the experiments performed with trap functions and show that ADMGA may improve GAs on some dynamic optimization scenarios. Robustness is also addressed and results show that ADMGA maintains a more stable performance over the wide range of dynamic scenarios. The congress HIS is mainly dedicated to hybrid models and real-world applications, so ADMGA was somewhat “lost” among other works. But good news came after the congress, and this line of research will probably make its way through other media.
P.S. In Brussels, avoid Hotel Continental, near Midi Station, unless you need inspiration for another insects’ heuristics.
Last friday I returned from the First Ferguson World Tour Of Science. I was traveling around Europe in order to present some things about my research.
First we arrived at Dortmund, Germany, where the Parallel Problem Solving from Nature (PPSN 2008) was celebrated. There I presented my paper “Evolving XSLT Stylesheets for Document Transformation” that you can download from here.
You can see me in this photo that JJ took explaining my poster to a great audience:
I hadn’t any problem explaining it. Nobody came to complain about colours, XML future, or typos in the word “Conclusions”. Oh, wait, I could remember… no, nobody came.
Here you can see me doing some magical science. Mmmh, who will be these guys sitting in that table?
After that I went to Castellón (Valencia, Spain) to present my paper “Algoritmos evolutivos distribuidos sobre dispositivos Bluetooth” (in Spanish, as you can see, but you also can download from here).
Just to say that it has been an awesome travel around the world! I met a lot of interesting people and learnt a lot. Yeah, I really love the researcher’s life.
The title of this post is due to today I received my MSc degree (I presented this paper in Granada, where I am a grad student).
