The TRI workshop series aims at investigating the existing and further synergies between the Multi-agent Systems (MAS), Machine Learning (ML) and Complex System (CS) disciplines. MAS can efficiently manage domains with distributed data and expertise. Also, they have the ability to solve large and complex problems. The expertise of each agent can focus on specific computational intelligence models such as learning classifiers, evolutionary algorithms, swarm intelligence techniques, or other specific optimization/learning algorithms. Applications range from bioinformatics and traffic networks to information retrieval and text classification. Broadly speaking, the area of CS is grounded on similar ideas: complex systems investigate how relationships between components of a system give rise to emerging collective behaviors and how the system as a whole and its components interact with an environment. ML has being used both in MAS as well as in CS when the agents or components of the system need to learn to make decisions. Moreover, since MAS and CS are becoming large and more and more complex, ML is key to improve performance. Here, not only the traditionally applied method of reinforcement learning shall be studied, but there is potential for using supervised and non-supervised ML techniques as well. Since these synergies among the three areas are not well studied, we see room for such a workshop where researchers and practitioners from the three areas could come together, as currently these three communities do not meet often.
The second edition of the TRI workshop is going to be held in July 27th and will be co-located with the International Joint Conference on Artificial Intelligence 2015 (IJCAI-15), which will be held in Buenos Aires, Argentina, from July 25th to 31st, 2015.