The main approaches for learning Bayesian networks can be classified as constraint-based, score-based or hybrid methods. Although high-dimensional consistency results are available for the constraint-based PC algorithm, such results have been lacking for score-based and hybrid methods, and most hybrid methods are not even proved to be consistent in the classical setting where the number of variables remains fixed.
Electroencephalography (EEG)-based brain-computer interfaces (BCIs) promise to provide a novel access channel for assistive technologies, including augmentative and alternative communication, and control of external devices for people with severe speech and physical impairments (SSPI). Research on the subject has been accelerating significantly in the last decade and the research community has taken great strides toward making EEG-based BCI a practical reality for individuals with SSPI.
Electronic health records (EHRs) now serve health care professionals in 95 percent of hospitals nationwide, a 9-fold increase from a decade ago.
The Public Health Dynamics Laboratory (PHDL) at the graduate school of Public Health at the University of Pittsburgh is a mathematical modeling group that uses computational methods to represent diseases and potential interventions. Initially funded through the Modeling Infectious Disease Agents Study (MIDAS), the PHDL has developed expertise in large scale, geospatially accurate agent-based modeling, primarily with a concentration on infectious diseases. These models require significant amounts of data, and we have championed he use of “synthetic” populations that statistically represent the real populations of interest. This talk will describe the ongoing modeling efforts at PHDL, our development of dynamic synthetic populations that age, marry, divorce and create social networks, and acquire non-infectious diseases, and describe our aspirational goals of the integration of individual data with complex, mechanistically-based models.
Many valuable datasets that could be used to counter epidemic threats are not used due to challenges in accessing and standardizing datasets, and in integrating data into novel analyses such as epidemic simulation. Our research aims to improve the acquisition, standardization, and integration of information about epidemic threats.
Internet support groups (ISGs) that enable individuals with similar conditions to assess and exchange self-help information and emotional support have proliferated in recent years. However, their benefit has not been established as randomized trials.
Microbial communities exist throughout the biosphere including the associations they form with humans, plants and animals. Understanding the diversity and genetic complexity of these communities, along with the interactions they undertake both within communities and with their environment has given rise to the concept of the study of the microbiome. With improvements in molecular biology, computational power, and high throughput technologies such as the advent of next generation sequencing, new opportunities exist to study the microbiome from multiple environments including their role in human health and disease. However, in order to maximize the information we can realize from these data types, careful considerations need to be made in terms of study design, and new tools and approaches are needed for data generation, analysis and interpretation. This talk will provide an overview of these topics using examples from microbiome studies.
In this talk, we will discuss some of the recent research on improving the performance of personalized recommender systems by using knowledge graphs (KG) to uncover the long range preferences of users.