• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Message from the EEML 2013 chairs

Global financial crisis of the recent years has made economists reconsider the path Economics as a discipline should take. A few examples illustrate the most vivid cases. Graeme Maxton published the book titled “The End of Progress: How Modern Economics Has Failed Us”, while New York Times hosted the round table “Rethinking How We Teach Economics”. Likewise, Paul Krugman declared: “ … the central cause of the profession’s failure was the desire for an all-encompassing, intellectually elegant approach that also gave economists a chance to show off their mathematical prowess. Unfortunately, this romanticized and sanitized vision of the economics led most economists to ignore all the things that can go wrong. They turned a blind eye to the limitations of human rationality … to the problems of institutions that run amok; to the imperfections of markets … and to the dangers created when regulators don’t believe in regulation.”

As any other science, Economics is undergoing a transition to the modern era, where innovations are achieved through a great number small steps taken by various research groups from all over the world. Thus, it should not be a science of single inventors, working with untested and unverified theories, but instead it should deal with regularities from the data, abundantly available nowadays. In particular, the way to design policies for poverty prevention is no longer a merely analytical consideration, but a case for state-of-the-art Data mining. It is common knowledge that people do not behave rationally and Data mining presents the most convenient way to understand how they actually behave and how the policy should respond to it. 

The time for economists to model people as rational actors is over. Data mining, Psychology and Neuroscience are about to become indispensable parts of Economics. The limitations of human reasoning have to be taken into account in order to prevent future crises. In the past, economists had little data and no computational power. Today, every interaction is recorded and computational power is abundant. That will make Economics a truly empirical science. 

As a part of renowned international conferences on various branches of machine learning, this full-day workshop intends to integrate scientists from Experimental Economics with those from AI. The first workshop – EEML 2012 – has been successfully conducted at ICFCA 2012 and we look forward to encouraging more and more interaction between researchers from both fields.

In Experimental Economics, laboratory and field experiments are carried out using human subjects in order to improve theoretical knowledge about human behavior during interaction. Although  financial rewards restrict subjects preferences in experiments, exclusive application of analytical game theory is not enough to explain the collected data. It calls for the development and evaluation of more sophisticated models. Additionally, the research area includes experiments, where human subjects are involved into interaction with automated agents. Nowadays experiments are conducted using state-of-the-art software like z-Tree, which produces massive text data sets.

The more data is used for evaluation, the more statistical significance can be achieved. Since large amounts of behavioral data are required to scan for regularities, along with automated agents needed to simulate and intervene in human interactions, Data Mining is the tool of choice for research in Experimental Economics.
 
This workshop is aimed at bringing together researchers from both Data Analysis and Economics in order to achieve mutually-beneficial results. We have received several good papers on the following topics: hotel market analyses in Europe by means of advanced regression models, stock market analyses by means of Machine Learning and Opinion Mining in Twitter, value leaks process mining by means of Formal Concept Analysis, distributed heterogeneous event mining and monitoring, and finding determinism in social learning in networks.

Rustam Tagiew
Dmitry I. Ignatov
Fadi Amroush