Christopher R. Stephens

Understanding the Complexities of Public Health through Data Mining

Resumen de la plática
Disease is a property of complex adaptive systems (CAS). Unfortunately, there does not currently exist an adequate theoretical framework within which to understand either complexity or adaptation. In the absence of theory per se, phenomenology and taxonomy become even more important. In the last few decades the data revolution has made available vast quantities of data associated with CAS and, in particular, disease and public health, with which we can potentially develop a much better understanding of the dynamics of CAS. However, the vast majority of this data is ""non-scientific"" or ""coincidental"", in the sense that it is not associated with sets of controlled scientific experiments designed to examine specific hypotheses. In this talk I will argue that data mining, as a phenomenological approach to understanding and modeling CAS, offers the most promise for developing a better understanding of them in practical settings such as public health and, indeed, in its automated form, is the only feasible way of analyzing the exponentially increasing amounts of data that are becoming available. I will illustrate these points using data sets associated with three important diseases: influenza(-like illness), type 2 diabetes and Leishmaniasis - an important emerging zoonosis. For influenza, I will discuss what we can learn from citizen participation systems, using a data mining of risk factors for ILI from the data associated with the Mexican REPORTA system. I will also discuss the difficulties of symptomatic diagnosis. For type 2 diabetes I will use analyses of large-scale public health surveys in Mexico to discuss risk factors for diabetes and the relationship between nutrition, obesity and diabetes, showing that quantity not quality is the most important factor for obesity. Finally, in the case of Leishmaniasis, I will show how large numbers of spatial-data sets can mined by converting them into complex inference networks which, in the case of species distributions, can then be used to predict reservoirs of Leishmaniasis or other emerging diseases."

Christopher Rhodes Stephens
C3 - Centro de Ciencia de la Complejidad, UNAM
México
 
Christopher Rhodes Stephens, nacio en Workington, Cumbria, Inglaterra, realizó sus estudios de licenciatura en Física en la Universidad de Queen´s College en Oxford de 1977-1980; realizó sus estudios de maestría y doctorado en Física, 1981-1984 y 1984-1986, respectivamente en la Universidad de Maryland en Estados Unidos. 
 
Sus campos de especialidad son: física estadística, Sistemas Complejos, Computación evolutiva y Minería de datos. Temas: grupo de renormalización, fenómenos colectivos, transiciones de fase, evolución, algoritmos genéticos, sistemas evolutivos, econofísica y finanzas, complejidad ecológica, enfermedades emergentes.
 
Ha sido árbitro en diversas revistas como Nature, Physics Reports, IEEE Transactions on Evolutionary Computation, Genetic Programming and Evolvable Hardware, Journal of Computing and Information Technology, Theoretical Computer Science, Artificial Life, Diversos conferencias en Evolutionary Computation, entre otras.
 
Ha impartido diversos cursos en Posgrado, tales como: mecánica cuántica relativista, física de muchos cuerpos, electrodinámica, teoría de campos, física de partículas elementales, fenómenos críticos y transiciones de fase, sistemas complejos, computación evolutiva, sistemas adaptables y algoritmos genéticos.
 
En experiencia industrial, Christopher Stephens es socio fundador de las empresas Adaptive Technologies SA de CV, una empresa de alta tecnología incubada a través de la Universidad Nacional Autónoma de México y Adaptive Technologies Inc. (Phoenix, AZ, EU).
 
Actualmente es investigador titular "C" de tiempo completo en el Instituto de Ciencias Nucleares de la UNAM. Es nivel III del Sistema Nacional de Investigadores.