21^st Mexican International Conference on Artificial Intelligence

The need of effective data flows for Machine Learning and Deep Learning; from Structure, Learning to Performance on Applications.

Abstract: Nowadays, the relevance of Machine Learning and Deep Learning has been increased in pattern recognition problems. In the last ten years of intensive research, with the popularity of deep learning and its capability to solve a huge variety of different problems, such as computer vision applications, increased the need to create better models under the paradigm of more hardware capabilities. In this talk, we will briefly review what aspects are needed in order to create and enhance solutions enabled by Machine Learning and Deep Learning, considering aspects of computing architecture, hardware-software codesign (i.e. GPU, optimization and inference platforms), as well as the orchestration of data for a batch or an online data-ingestion. We briefly show some examples for face recognition, identity recognition, digital twins and activity detection, for which, not only precision is necessary, but also affordability and sustainability for technical, ethical and environmental requirements.

Bio: Miguel González Mendoza holds a PhD degree and a Postdoc in Artificial Intelligence from INSA and LAAS-CNRS Toulouse, France, in 2003 and 2004 respectively. Since 2004 he works as research professor at Tecnologico de Monterrey, Mexico. Miguel González Mendoza’s research activities are focused on machine learning and deep learning, computer vision, areas in which has supervised 11 PhD and 28 MSc. Theses, published more than 130 peer reviewed scientific publications, participated and conducted more than 20 national (CONACYT founded) and international (European founded) research and innovation projects, and chaired 4 international Congresses, and had served as senior consultant in the Financial Sector. President of the Mexican Society for Artificial Intelligence (2017-2018), member of the Mexican National Research System (SNI) rank II, member since 2006. Member of the Mexican Academy of Sciences. Head of the Graduate Programs on Computer Sciences at Tecnologico de Monterrey, Mexico City 2005-2016. Young Scientist at the World Economic Forum for New Champions, 2012.

Combining Microservices, REST, Linked Data and Multi-Agent Systems

Abstract: The emergence of the Microservices and REST architectural styles within industry present an opportunity to rethink how Multi-Agent Systems (MAS) technologies can be integrated into mainstream software architecture. This talk will introduce a novel MAS integration strategy, known as Multi-Agent MicroServices (MAMS) that enables agents to be immersed seamlessly within Microservices architectures. It will explore the use of Semantic Microservices that expose Linked Data representations to implement decentralised hypermedia data structures that can easily be transformed into knowledge graphs that can be traversed by MAMS agents. The potential of the approach is illustrated through ongoing case studies in the areas of Agent-Based Transport Simulation and Digital Twins for Smart Agriculture.

Bio: Rem Collier is an Associate Professor in the UCD School of Computer Science, Dublin, Ireland, where he has been faculty since 2004. His primary research interests are focused on Multi-Agent Systems and in particular Agent-Oriented Software Engineering / Programming. From 1996 to 2010, he developed the Agent Factory Framework, a cohesive framework for the rapid prototyping of Multi-Agent Systems (MAS). During that time, he was involved in a range of projects exploring its application to Social Robotics, Virtual Reality and Mobile Computing. He was a post-doctoral researcher on the team that won the prestigious 2003 CIA Systems Innovation Award for work on the Agents Channelling ContExt Sensitive Services (ACCESS) architecture. Since 2017, his research interests have centred around the areas of MAS, Microservices, REST and Linked Data where he has proposed the Multi-Agent MicroServices (MAMS) architectural style as a novel approach to integrating agent technologies into Microservice-based software architectures. He has been Primary Supervisor of 10 PhD and 12 MSc theses and published over 120 academic papers. He is currently Lead-PI of CONSUS, a €16.5M nationally funded Strategic Partnership Project focussed on Digital Precision Agriculture & Crop Science and is a Principal Investigator of CAMEO a Disruptive Technologies Innovation Project charged with developing Ireland future Earth Observation platform.

Machine Learning and Big Data Analytics for Sustainable Food-Energy-Water Nexus

Abstract: Food, energy, and water (FEW) are the physical resources and natural systems in the earth for humanity surviving, sustainable food, energy and water have been seriously considered as the United Nations' Sustainable Development Goals 2030. In the last decade, huge amount of FEW data haven been accumulated thanks to the technology advances. Additionally, the FEW data are big, spatial, and dynamic, presenting the characteristics of Big Data such as Velocity, Variety, Veracity, Vulnerability, Viscosity, Vararity, etc. Data scientists have noticed the big opportunities in FEW nexus challenges.

In this talk, data science related problems in the FEW nexus, such as wind energy production prediction, land irrigation efficiency analysis, water quality monitoring, plant disease detection, precision agriculture, etc. will be addressed. These problems can be solved by machine learning and Big Data analytics techniques, corresponding topics are spatio-temporal data, high dimensionality, imbalance data, complexity of heterogeneous data, data fusion, etc.

Bio: Xiaoou Li has been a professor of Computer Science Department, at The Research and Advanced Studies Centre of the National Polytechnic Institute (CINVESTAV-IPN) in Mexico since 2000. She was a lecturer at Northeastern University from 1995 to 1997; then, she was a postdoc of the National Autonomous University of Mexico (UNAM) from 1998 to 2000. She was a senior research fellow of the School of Electronics, Electrical Engineering & Computer Science in Queen's University Belfast, UK from 2006 to 2007; and a visiting professor at the School of Engineering of the University of California Santa Cruz and Shenzhen University in 2010 and 2019 respectively. Her research interests include machine learning and data mining applications, social network analysis, knowledge-based system, Petri nets, neural networks, system modeling and simulation, smart manufacturing, human machine interface (HMI), educational robotics, etc. She has published more than 100 papers on international journals, books, and conferences. As PI, she has successfully finished three CONACYT (like NSF in Mexico) Basic Science projects in the field of Knowledge and Data Engineering, one collaborative project with the University of California Riverside. She is an active organizer (General Chair or Co-chair) of IEEE international conferences such as ICNSC, SMC, CASE, etc. She has been in Editorial Board of IEEE Press, associate editor of IEEE Transactions on Automation Science and Engineering, IEEE Access, IEEE/CAA Journal of Automatica Sinica, and IEEE SMC Magazine, Engineering Applications of Artificial Intelligence, etc. She is a regular member of the AMC (Mexican Association of Science) since 2008, and a member of SNI (National Researcher System of Mexico) level 2 .