Data Science in Screening Complex Systems with Numerous Factors

12.00 a 13.00
ITAM, Río Hondo
Salón 101
Conferencias y Seminarios
Data Science in Screening Complex Systems with Numerous Factors

Abraham Aldaco Gastelum, Ph.D.
Project Engineer Leader
Aviation Systems
General Electric

Complex systems are pervasive in science and engineering. Some examples include complex engineered networks such as the internet, the power grid and transportation networks. The understanding of such complex systems is limited because their behavior cannot be characterized using traditional techniques of modelling and analysis. Typically, a domain-expert, based on his/her knowledge and experience, can determine which are the variables or factors more significant in just one part of a complex system. However, it is highly questionable that a domain-expert can determine which are the most important factors in a complex system as a whole. Much more questionable is that the domain-expert can determine the most important interactions of the factors for the entire system or even for part of it. Thus, if assumptions have been taken into account to remove factors a priori, conducting an empirical or simulated experiment, as well as its results, is of unreliable accuracy and precision. Moreover, traditional approaches for screening are ineffective for complex systems because of the size of the experimental design. To address this problem, a combinatorial design is used as a screening design for complex systems. Combinatorial arrays exhibit logarithmic growth in the number of factors.

División Académica de Ingeniería


Mantente en Contacto
Todos los derechos reservados © ITAM, 2018. Río Hondo No.1, Col. Progreso Tizapán, Álvaro Obregón, 01080 Ciudad de México, México, Tel. +52(55) 5628 4000