A5011520392- Scientific Computing and Data Management
- Distributed and Parallel Computing Systems
- Research Data Management Practices
- Cloud Computing and Resource Management
- Distributed systems and fault tolerance
- Data Quality and Management
- Advanced Data Storage Technologies
- Parallel Computing and Optimization Techniques
- Peer-to-Peer Network Technologies
- Semantic Web and Ontologies
- Reinforcement Learning in Robotics
- Smart Grid Energy Management
- Neuroscience and Neuropharmacology Research
- Zebrafish Biomedical Research Applications
- Advanced Database Systems and Queries
- Neuroinflammation and Neurodegeneration Mechanisms
- Service-Oriented Architecture and Web Services
- Genetics, Bioinformatics, and Biomedical Research
- Big Data and Business Intelligence
- Biomedical Text Mining and Ontologies
University of Southern California
1998-2022
Marina Del Rey Hospital
2020
LAC+USC Medical Center
2015
Southern California University for Professional Studies
1998-2005
Southern States University
2004
University of Illinois Urbana-Champaign
2002
University of Illinois Chicago
2002
Grid technologies enable large-scale sharing of resources within formal or informal consortia individuals and/or institutions: what are sometimes called virtual organizations. In these settings, the discovery, characterization, and monitoring resources, services, computations challenging problems due to considerable diversity; large numbers, dynamic behavior, geographical distribution entities in which a user might be interested. Consequently, information services vital part any software...
Grid computing is concerned with the sharing and coordinated use of diverse resources in distributed "virtual organizations." The dynamic multiinstitutional nature these environments introduces challenging security issues that demand new technical approaches. In particular, one must deal local mechanisms, support creation services, enable trust domains. We describe how are addressed two generations Globus Toolkit/spl reg/. First, we review Toolkit version 2 (GT2) approach; then approaches...
Applications designed to execute on "computational grids" frequently require the simultaneous co-allocation of multiple resources in order meet performance requirements. For example, several computers and network elements may be required achieve real-time reconstruction experimental data, while a large numerical simulation access supercomputers. Motivated by these concerns, we have developed general resource management architecture for Grid environments, which is an integral component. We...
We often encounter in distributed systems the need to model, access, and manage state. This state may be, for example, data a purchase order, service level agreements representing resource availability, or current load on computer. introduce two closely related approaches modeling manipulating within Web services (WS) framework: Open Grid Services Infrastructure (OGSI) WS-Resource Framework (WSRF). Both define conventions use of definition language schema that enable management OGSI...
In this paper we study a minimalist decentralized algorithm for resource allocation in simplified Grid-like environment. We consider system consisting of large number heterogenous reinforcement learning agents that share common resources their computational needs. There is no communication between the agents: only information receive (expected) completion time job it submitted to particular and which serves as signal agent. The results our experiments suggest can be used improve quality scale system.
One of the criteria for Grid infrastructure is ability to share resources with nontrivial qualities service. However, sharing in Grids complicated that requires bridge differing policy requirements resource owners create a consistent cross-organizational domain delivers necessary capability end user while respecting owner. Further complicating management need coordinate usage, diversity types and variety different modes may be used. We present unifying framework which we can address these...
Defining the structural and functional changes in nervous system underlying learning memory represents a major challenge for modern neuroscience. Although neuronal activity following formation have been studied [B. F. Grewe et al.,
The development of applications and tools for high-performance "computational grids" is complicated by the heterogeneity frequently dynamic behavior underlying resources; complexity themselves, which often combine aspects supercomputing distributed computing; need to achieve high levels performance. Globus toolkit has been developed with goal simplifying this application task, providing implementations various core services deemed essential computing. In paper, we describe two large toolkit:...
The Common Fund Data Ecosystem (CFDE) has created a flexible system of data federation that enables researchers to discover datasets from across the US National Institutes Health without requiring owners move, reformat, or rehost those data. This is centered on catalog integrates detailed descriptions biomedical individual Programs' Coordination Centers (DCCs) into uniform metadata model can then be indexed and searched centralized portal. Crosscut Metadata Model (C2M2) supports wide variety...
The pace of discovery in eScience is increasingly dependent on a scientist's ability to acquire, curate, integrate, analyze, and share large diverse collections data. It all too common for investigators spend inordinate amounts time developing ad hoc procedures manage their In previous work, we presented DERIVA, Scientific Asset Management System, designed accelerate data driven discovery. this paper, report the use DERIVA number substantial applications. We describe lessons have learned,...
The overhead and burden of managing data in complex discovery processes involving experimental protocols with numerous data-producing computational steps has become the gating factor that determines pace discovery. lack comprehensive systems to capture, manage, organize retrieve throughout life cycle leads significant overheads on scientists' time effort, reduced productivity, reproducibility, an absence sharing. In "creative fields" like digital photography music, asset management (DAM) for...
Increasingly, scientific discovery is driven by the analysis, manipulation, organization, annotation, sharing, and reuse of high-value data. While great attention has been given to specifics analyzing mining data, we find that there are almost no tools nor systematic infrastructure facilitate process from We argue a more perspective required, in particular, propose data-centric approach which stands on foundation data collections, rather than fleeting transformations operations. To address...
Computational grids are enabling collaboration between scientists and organizations to generate archive extremely large datasets across shared, distributed resources. There is a need visually explore such data throughout the life-cycle of projects. Practical exploration requires visualization tools that can function in same grid environment which created stored. Resource management interfaces an important structural component computing environments because they enable uniform access wide...
Database evolution is a notoriously difficult task, and it exacerbated by the necessity to evolve database-dependent applications. As science becomes increasingly dependent on sophisticated data management, need an array of database-driven systems will only intensify. In this paper, we present architecture for data-centric ecosystems that allows components seamlessly co-evolve centralizing models mappings at service pushing model-adaptive interactions database clients. Boundary objects fill...
The foundation of data oriented scientific collaboration is the ability for participants to find, access and reuse created during course an investigation, what has been referred as FAIR principles. In this paper, we describe ERMrest, a collaborative management service that promotes by enabling throughout life cycle. ERMrest RESTful web discovery organizing diverse assets into dynamic entity relationship model. We present details on design implementation its performance use range...
Biomedical research depends upon increasingly high throughput instruments and sophisticated data analytics. In spite of the significant overhead handling data, there is little support for researchers to manage organize purposes exploration, analysis, ultimately publication. Shared file systems with metadata coded into directory hierarchies spreadsheets are common practice. this paper, we present a digital asset management approach system streamlining operations reducing overheads biomedical...
Abstract The Common Fund Data Ecosystem (CFDE) has created a flexible system of data federation that enables users to discover datasets from across the U.S. National Institutes Health without requiring owners move, reformat, or rehost those data. CFDE’s is centered on catalog ingests metadata individual Program’s Coordination Centers (DCCs) into uniform model can then be indexed and searched centralized portal. This Crosscut Metadata Model (C2M2) supports wide variety types terms used by...
Creating and maintaining an accurate description of data assets the relationships between is a critical aspect making findable, accessible, interoperable, reusable (FAIR). Typically, such metadata are created maintained in catalog by curator as part publication. However, allowing to be producers generated rather then waiting for publication can have significant advantages terms productivity repeatability. The responsibilities management need not fall on any one individual, but may delegated...