The focus of the school will be on large-scale data analytics, which lies at the intersections of data analytics algorithms and high performance computing systems.
Students will be introduced to problems in data analytics arising from both the machine learning community and the scientic computing community. The school will also include perspectives from industry, such as Hodge Star Scientific Computing and IBM, as well as from academic instructors. A unique strength of this school will be to expose students to “end-to-end” multidisciplinary topics, which span several traditionally disparate areas. The courses will give a background on methods and algorithms for data analytics, approximation algorithms to deal with large volumes of data, languages and tools for implementing those algorithms on large scale computers, and data-driven applications from scientific computing and machine learning.
(If the program doesn’t appear, please click below)
- High-dimensional data analysis and underlying algebraic problems, Haesun Park
- Approximation algorithms for high dimensional problems : tensors and their low rank approximations, Tammy Kolda
- Introduction to Analysis on Symmetric Cones, Jack Poulson
- Languages and tools, Costas Bekas
The lectures will also highlight directions of research in this area, emphasize already existing as well as potential synergies between data analytics and high performance computing, and outline prospects for the future.
The school brings together two different communities, namely, high-performance computing and machine learning/data mining. The students will have a broad exposure to a mix of applied mathematics, software, and scalable HPC and cloud systems. The school also brings a strong industrial perspective.
Haesun Park, Georgia Institute of Technology, USA.
Tammy Kolda, Sandia National Lab.
Jack Poulson, Research Scientist, Hodge Star Scientific Computing.
Costas Bekas, IBM Zurich.