Dr. Yoon’s project on “Objective-Driven Data Reduction for Scientific Workflows” recently funded by the U.S. Department of Energy (DOE), Advanced Scientific Computing Research (ASCR), has been featured in a recent article posted on the Texas A&M Department of Electrical and Computer Engineering website.
The full article can be accessed at the link below:
The following news release was issued by the U.S. Department of Energy. It announces funding for nine projects that will address management and processing of massive data sets produced by scientific observatories, experimental facilities, and supercomputers that span the DOE national laboratory complex.
As part of this program, Brookhaven Lab was awarded $2.4 million in funding over three years. Dr. Byung-Jun Yoon of Brookhaven National Lab‘s Computational Science Initiative will lead the project with Texas A&M University and University of Illinois Urbana-Champaign as partners. Dr. Yoon and his team’s work will involve using novel theoretical strategies to develop practical algorithms anchored in scientific objectives – i.e., focused specifically on the goals of interest. These algorithms would bypass scientifically irrelevant information, which could considerably reduce overall data generated by complex experimental or computational systems and streamline any required processing.
The seventh annual New York Scientific Data Summit, NYSDS 2021, is a multi-day interactive virtual event that will focus on Applied Mathematics, HPC, and AI challenges in diverse areas, such as climate science, medical research, critical infrastructure and other cross-cutting research topics.
Register online now at https://www.bnl.gov/nysds21/reg/step1.php. There is no registration fee, but registration is required for attendance. Registration deadline is Thursday, October 21, 2021. The virtual event will be held on Tuesday, October 26 through Friday, October 29, 2021.
NYSDS will provide a great opportunity to connect with researchers, developers, and end-users to inform efforts that will drive data research forward with new ideas and partnerships.
Dr. Yoon is co-organizing a Data Hackathon event at the 2021 IEEE Healthcare Summit (IHS). The Data Hackathon focuses on COVID-19 Datasets with the goal of demonstrating the potentials of Biomedical Health Informatics (BHI) and Artificial Intelligence (AI) to combat pandemics.
The Data Hackathon features a wide range of datasets and challenges, which include:
Individuals or teams interested in participating in the Hackathon should register by Sep. 15, 2021 and submit their report and code by Sep. 26, 2021. Many awards sponsored by IEEE and other sponsoring institutions will be given to the top performing participants.
We are happy to announce that a Data Hackathon event will be organized at the First IEEE Healthcare Summit (IHS) – a virtual event that will be held on Oct. 4-7, 2021.
The main focus of the Data Hackathon will be on integrating the latest advances in Biomedical Health Informatics and AI (artificial intelligence) to combat COVID-19 pandemics. Multiple challenges based on various datasets – including single cell RNA sequencing data, CT imaging data, audio data, public health data, NLP (natural language processing) data, and others – will be provided, where participants can choose the challenge and dataset based on their interest and expertise.
Teams/individuals interested in participating in the Data Hackathon should register by the registration deadline (Sep. 15, 2021). Many awards – sponsored by IEEE as well as other individual institutions sponsoring the event – will be given to top performing teams/individuals.
The first annual IEEE Healthcare Summit will be a virtual event held on 4-7 October 2021. This event aims to report progress made in the fight against COVID-19, and in better preparations for future pandemics, through the integration of Artificial Intelligence (AI) methods and tools with Biomedical and Health Informatics (BHI):
Translational bioinformatics: use genomics and proteomics to do SARS-CoV-2 subtyping, and build tools to develop targeted vaccine or drug, to do early screening to limit outbreak, and to perform evolution trajectory prediction;
Sensor informatics: use real-time wearable sensor data to monitor asymptomatic and mild-symptom home-based patients, to treat severe-symptom patients in hospitals, and COVID-related sequelae (e.g. neurological and cardiac disease);
Imaging informatics: use CT, X-rays, MRI, ultrasound, and other imaging modalities with RT-PCR data to improve diagnosis, prognosis, and monitoring of patients with COVID-19, or other infectious diseases in the event of future pandemics;
Clinical informatics: use multimodal data to find effective clinical care workflows for critically ill patients, to track the occurrence of COVID-related sequelae for long term follow-up, and to perform risk assessment and decision making;
Public health informatics: use epidemiological models to analyze outbreak data for supporting population health management, resource supply chains, and future care preparation;
Mental health informatics: collect and analyze mental health data during the pandemic to model behavior changes caused by the pandemic, and to evaluate the consequences of policies for future preparedness;
Mixed VR/AR and Robotics: integrate BHI with VR, AR, and robotics to effectively visualize omic, imaging, sensor, and population pandemic data, to train medical robots, and to assist public health policy making for preparedness; and
Fairness and ethics: use pandemic data to identify regional, racial, and ethnic disparities in infection and vaccination rates and the causal factors of the disparities, for the preparedness of future pandemics and fair resource allocation.
The 2nd Workshop on Knowledge Guided Machine Learning (KGML2021) is inviting researchers working in relevant fields to share their latest research findings, including work in progress.
KGML2021 is a workshop that brings together data scientists – researchers in data mining, machine learning, and statistics – and researchers from hydrology, atmospheric science, aquatic sciences, and translational biology to discuss challenges, opportunities, and early progress in bringing scientific knowledge to machine learning.
KGML2021 is a virtual conference from August 9-11 and welcomes submissions from researchers around the world.
The ACM SIGBio bridges computer science, mathematics, statistics with biology and biomedicine. The mission of ACM SIGBio is to improve our ability to develop advanced research, training, and outreach in Bioinformatics, Computational Biology, and Biomedical Informatics by stimulating interactions among researchers, educators and practitioners from related multi-disciplinary fields.
Dr. Yoon’s term as the Treasurer will be from July 1, 2021 until June 30, 2024. For more information about ACM SIGBio, please visit http://www.sigbio.org
Various real-world applications involve modeling complex systems with immense uncertainty and optimizing multiple objectives based on the uncertain model. Being able to quantify the impact of such model uncertainty on the operational objectives of interest is critical, for example, to design optimal experiments that can most effectively reduce the uncertainty that affect the objectives pertinent to the application at hand. In fact, such objective-based uncertainty quantification (objective-UQ) has been shown to be much more efficient for optimal experimental design (OED) compared to other approaches that do not explicitly aim at reducing the “uncertainty that actually matters”.
The concept of MOCU (mean objective cost of uncertainty) provides an effective means to quantify this objective uncertainty, but its original definition was limited to the case of single objective operations.
In our recent paper, we extend the original MOCU to propose the mean multi-objective cost of uncertainty (multi-objective MOCU), which can be used for objective-based quantification of uncertainty for complex uncertain systems considering multiple operational objectives:
Based on several examples, we illustrate the concept of multi-objective MOCU and demonstrate its efficacy in quantifying the operational impact of model uncertainty when there are multiple, possibly competing, objectives.
For more information regarding the concept of objective-UQ, optimal experimental design (OED), and other relevant resources, please visit: https://objectiveUQ.org
The multi-objective MOCU quantifies the expected performance gap between the robust multi-objective operator that needs to be used to main good performance in the presence of model uncertainty and the optimal multi-objective operator for the true (but unknown) model.
The final version of our paper entitled “Physics-constrained Automatic Feature Engineering for Predictive Modeling in Materials Science” recently presented at the 35th AAAI Conference on Artificial Intelligence (AAAI-21) is now accessible at the link below:
Automatic Feature Engineering (AFE) aims to extract useful knowledge for interpretable predictions given data for the machine learning tasks of interest. In this work, we presented a novel AFE scheme that effectively extracts relationships from data that can be interpreted based on functional formulas to discover their “physical meaning” or “new hypotheses”. Here we focused on materials science applications, where interpretable predictive modeling may enhance our understanding of materials systems and also guide the discovery of new materials.
Typically, it is computationally prohibitive to exhaustively explore all potential relationships to identify interpretable and predictive features. We overcome this challenge by designing an AFE strategy that efficiently explores a feature generation tree (FGT) using a deep Q-network (DQN) for scalable and efficient exploration of the feature space in an automated manner. We demonstrate that our proposed DQN-based AFE strategy yields promising results when benchmarked against existing AFE methods based on several materials science datasets.