These are some of the projects this research area is currently participating in:
Active Recommendation Project Development of recommendation systems for large databases and the WWW, which adapt to the expectations of users. Recommendation applications are based on biological metaphors, to be reactive, adaptive, and evolving to and with their user communities. We also employ Network Analysis methods to discover Knowledge and produce Recommendations. Current recommendation algorithms are implemented for the MyLibray Portal at Los Alamos.
Design Principles of Genetic Regulatory Networks (LDRD-ER). To explain the diversity of regulatory structures of gene circuits in nature, we are comparing model networks according to evolutionary criteria, and seeing whether there are classes of models that may confer evolutionary advantage. We have been applying such a procedure to study design principles of gene regulation, using the method of Mathematically Controlled Comparison developed by Michael Savageau.
Identification of Interests, Trends and Dynamics in Document Networks Project (LDRD-ER) Research on of metric properties of document networks extracted from co-occurence data. Semi-metric edges in distance graphs have been characterized as containing latent information which is indicative of relevance to user communities, and reveals trends in the dynamics of these networks.
Improving Local Search (LDR-ER)Using techniques and models from sources as diverse as graph theory, computational complexity and statistical physics, we explore the performance of Extremal Optimization and relate this to the behavior of other local search methods.
Knowledge, culture and the structure of human society. We are testing the hypothesis that dynamical processes set much of the structure of social networks. To achieve this we are applying and developing network analysis methods to paradigmatic examples of scientific discovery. We test these methods against available historic knowledge.
Marketecture - Modeling and Simulation of Deregulated Electricity Market
Protein Function Inference (LDRD-DR) Development of methods for inferring protein function based on analysis of sequence, structure, text information sources, and biological databases
Statistical Physics of Infrastructure Networks. Investigation of universal principles governing the evolution and dynamics of infrastructure networks, such as transport
infrastructures, social interactions, communications networks, protein
networks, and the Internet.
Recent Projects
These are some of the projects this research area has recently participated in:
(Gene) Expression Array Analysis as part of the Functional Genomics Research Program from the Center for Human Genome Studies at the Los Alamos National Laboratory. Investigation of advanced computational and statistical techniques to integrate genetic information in several databases to predict protein function.
Extremal Optimization
. The goal is to develop general-purpose methods for finding high-quality solutions to hard optimization problems, inspired by self-organizing processes often found in nature.
Combinatorial optimization in biology. This project addresses theoretical and practical aspects of
optimization problems arising in a biologically-motivated setting.
Major efforts include algorithms for optimizing the DNA sequencing
process, and pooling designs for maximally efficient group testing.
Agent-based modeling of socio-technical organizations . The project aimed to develop formal models and agent-based simulations of socio-technical organizations conceived as hybrid systems with dynamical components (e.g. a physical environment) in interaction with decision-making agents empowered with knowledge structures.