CAL-R

We have released a simulator of Temporal Difference that works with Complete Serial Compounds (CSC). It is able to run whole forward, backward and simultaneous conditioning designs with fixed and variable length CSs. It allows the user to work with different contexts and with stimulus and context-stimulus compounds, configural cues and different eligibility traces. The user can also modify the US parameters from phase to phase.

The R&W Simulator ver. 3.1 has been released. This version corrects a bug found in the figure legends that showed unnecesary extra markers for all combinations of stimuli and groups. It also sets a smaller beta- value by default.

New paper A Java simulator of Rescorla and Wagner's prediction error model and configural cue extensions to appear in Computer Methods and Programs in Biomedicine.

The Centre for Computational and Animal Learning Research aims at encouraging interdisciplinary research in learning and behaviour by strengthening collaboration between learning theorists, biologists, neuroscientists, cognitive scientists, mathematicians, software developers and computer scientists. In order to do so, the Centre promotes the distribution of ideas, results and tools through its publications and the organization of workshops where researchers can debate about the area's state of the art and future trends. Eduardo Alonso and Esther Mondragón are the Centre's founders and co-directors.

The main activities we are involved in are:

We develop simulators of psychological models of classical conditioning:

We have launched a Rescorla & Wagner Simulator that allows to calculate compound stimulus' associative strength with and without configural cues, produces numerical and graphical outputs, and exports the data to ".xlsx" files.
We have just released a Temporal Difference simulator that is able to run trial-based learning designs and allows the user to use different contexts, stimulus and context-stimulus compounds, configural cues and different eligibility traces (accumulating trace, bounded accumulating trace and replacing trace).

We have edited a book entitled Computational Neuroscience for Advancing Artificial Intelligence: Models, Methods and Applications with IGI Global publishers.

The book "...captures the latest research in this area, providing a learning theorists with a mathematically sound framework within which evaluate their models. The significance of this book lies in its theoretical advances, which are grounded in an understanding of computational and biological learning. The approach taken moves the entire field closer to a watershed moment of learning models, through the interaction of computer science, psychology and neurobiology."

Also available at

We are planning to organise a workshop on Associative Learning and Reinforcement Learning --a sequel of an event we held in 2006 as part of the AISB annual convention.

Finally, The Centre is launching a publication, the CAL Archive. This on-line publication will serve as a quarterly free repository of theoretical and experimental results including null results and preliminary work. Our editorial approach is focused on the content rather than cumbersome formatting, hence instructions for authors are to be kept to a minimum. We plan to send the first call for papers on June 2012.