MRI: Acquisition of Big-Data Private-Cloud Research Cyberinfrastructure (BDPC)

提案编号：13- 38099 PI：Vardi，Moshe;艾伦，Genevera; Bradshaw，Stephen; Karaki，Lydia; Veeraraghavan，Ashok机构： Rice University职称：MRI/采集：大数据私有云研究网络基础设施（BDPC） 项目建议：该项目获得了一种用于大数据云计算的新型网络基础设施工具，该工具被设计为具有大内存需求的松散耦合计算系统，可以实现大量的应用领域以及对云计算基础设施的研究。领域科学涉及的范围从大数据软件技术的开发，跨越基础计算机科学工作，到电子医疗记录，twitter流和飓风疏散策略的分析。在理解疾病和治疗方法，以及在机器学习，优化，压缩传感，图像处理，统计分析和数据挖掘领域的数学模型的开发和应用方面，预计会有额外的好处。该仪器还将有助于弥合天体物理学数值模型与观测之间的差距。更广泛的影响：对社会的更广泛影响，特别是在教育和培训方面（包括对代表性不足的群体的成员）都是引人注目的。该仪器将直接影响所有学生在计算和计算问题解决课程的教育经验。所针对的研究界是多样和广泛的，包括代表性不足的群体，具有很强的经验和实验成分。所提出的仪器非常适合于培训和教育。

Proposal #: 13-38099PI(s): Vardi, Moshe; Allen, Genevera; Bradshaw, Stephen; Karaki, Lydia; Veeraraghavan, AshokInstitution: Rice UniversityTitle: MRI/Acq.: Big-Data Private Cloud Research Cyberinfrastructure (BDPC) Project Proposed:This project, acquiring a novel cyber-infrastructure instrument for big data cloud computing designed as a loosely coupled computations system with large memory requirements, enables a significant range of application domains as well as research into infrastructures for cloud computing. The domain sciences addressed range from development of big-data enabling software technologies spanning fundamental computer science work, to the analysis of electronic medical records, twitter streams, and hurricane evacuation strategies. Additional benefits are expected in understanding disease and therapeutic treatments, and in the development and application of mathematical models in the areas of machine learning, optimization, compressed sensing, image processing, and statistical analysis and data mining. The instrument will also help bridge the gap between numerical models and observations in astrophysics.Broader Impacts: The broader impacts on society, and especially in education and training (including for members of underrepresented groups) are all compelling. The instrument will directly impact the educational experience for all students taking classes in computing and computational problem solving. The targeted research communities are diverse and broad, including the underrepresented groups, with strong empirical and experimental components. The proposed instrument is highly suitable for training and education.