York Centre For Complex Systems Analysis - YCCSA

Look out for YCCSA's Sam Ellis, talking about wood ants on the One Show on Thursday 24 Oct.

Seminar: Karoline Wiesner: Thermodynamics of complexity

Friday, 25 October 2013, 13:30

Where: Univ. of York, Ron Cooke Hub: RCH/204

http://www.maths.bris.ac.uk/~enxkw/K_Wiesner_Home.html

"Physical systems are often simulated using a stochastic computation where different final states result from identical initial states. Here, I derive the minimum energy cost of simulating a data sequence of a general physical system by stochastic computation. Using information theory, I show that the cost is proportional to the difference between two information- theoretic measures of complexity of the data—the statistical complexity and the predictive information. I derive the difference as the amount of information erased during the computation. To this framework I will add a quantum mechanical measure of complexity. The resulting triple opens an interesting view on thermodynamics, quantum mechanics and complexity."

Relevant papers

K Wiesner, M Gu, E Rieper, and V Vedral. “Information-theoretic Lower Bound on Energy Cost of Stochastic Computation.” Proceedings of the Royal Society A: Mathematical, Physical and Engineering Science 468, no. 2148 (December 8, 2012): 4058–4066. doi:10.1098/rspa.2012.0173.

M Gu, K Wiesner, E Rieper, and V Vedral. “Quantum Mechanics Can Reduce the Complexity of Classical Models.” Nature Communications 3 (March 27, 2012): 762. doi:10.1038/ncomms1761.

Karoline Wiesner

We had a great time at the Scone Seminar today! Special thanks to Grazziela Figueredo for her fascinating talk on Agent Based Simulation.

Seminar Friday, 18 October 2013

Time: 13:30

Where: Univ. of York, Ron Cooke Hub: RCH/204

http://www.nottingham.ac.uk/mathematics/people/grazziela.figueredo

Early-stage cancer and its interactions with the immune system are still not fully understood. In order to better understand these processes, researchers employ different methods. Simulation and in particular, agent-based simulation (ABS) have been found useful tools for understanding it.

In a previous study we have built an ABS model to study the interplay of immune cells and early-stage cancer. The model considers interactions between tumour cells and immune effector cells, as well as the immune-stimulatory and suppressive cytokines IL-2 and TGF-Beta. IL-2 molecules mediate the immune response towards tumour cells. They interfere on the proliferation of effector cells according to the number of tumour cells in the system. Conversely, TGF-Beta stimulates tumour growth and suppresses the immune responses by inhibiting the activation of effector cells and reducing tumour-antigen expression.

In order to validate our model, we used a well-established mathematical model found in the literature. While at average both models do not show a statistical significant difference, some additional trends in the results of the ABS model are observed. As ABS is a stochastic simulation method, it was run for multiple times. Instead of having one solution, as it is the case for a deterministic mathematical model, ABS produces a variety of outcomes. These solutions are usually very similar. In our cases study, however, we could observe some instances which could not have been observed by using analytical methods).

The use of ABS modelling has therefore led to the discovery of additional “rare” patterns, which we would have not been able to derive by using analytical methods. These “extreme cases” indicate that there might be circumstances where the tumour cells are completely eliminated by the immune system, without the need of any cancer therapies. We strongly believe that the observed emergent behaviour produced by stochastic simulation can make a useful contribution to assisting immunological research. With the additional information supplied from the ABS, immunologists can test new hypotheses and further investigate whether these extreme cases actually occur in reality and why.

Grazziela Figueredo - The University of Nottingham

Peter Cowling, Ed Powley and Daniel Whitehouse of YCCSA will co-present a paper with collaborator Jeff Rollason of AI Factory Ltd at the AIIDE13 conference (www.aiide.org) on integrating AI search into a leading commercial game with millions of downloads.

aiide2013 AIIDE is the definitive point of interaction between entertainment software developers interested in AI and academic and industrial AI researchers. Sponsored by the Association for the Advancement of Artificial Intelligence (AAAI), the conference is targeted at both the research and commercial commu...

Come along to our Scone Seminar, this Friday at 13:30 in RCH/204. Our guest speaker will be Dr. Grazziela Figueredo from the University of Nottingham. For more information, visit http://www.york.ac.uk/yccsa/activities/news-events/

Events - YCCSA, The University of York Guidance for Seminar SpeakersThe YCCSA Scone Seminars cover the theme of complexity in its broadest sense and are an ideal opportunity to hear about different research perspectives, study systems and analytical approaches. The audience is diverse, from PhD students to professors, including represent...