ASE International Conference on Big Data Science 2014: Day 2 Highlights

Highlights from the presentations by Data Science leaders from USC, YarcData and Revolution Analytics on day 2 of ASE Conference on Big Data Science 2014 held in Stanford University.

The Second ASE International Conference on Big Data Science was a great opportunity for students, data scientists, engineers, data analysts, and marketing professionals ASEto learn more about the applications of Big Data. Session topics included “Enabling Science from Big Image Data,” “Engineering Cyber Security and Resilience,” “Cloud Forensics,” and “Exploiting Big Data in Commerce and Finance.”

Held at the Tresidder Memorial Union at Stanford University, the ASE International Conference on Big Data Science took place from Tuesday, May 27 – Friday, May 31, 2014.

Highlights from workshops.

Highlights from Day 1.

Here are highlights from Day 2 (Thursday, May 29, 2014):

Kai HwangDr. Kai Hwang, Professor, University of Southern California kicked off second day with a talk on “Enabling Cloud Analytics for Big-Data Security and Intelligence”. He addressed the growing interest in big-data science surrounding the use of cloud analytics, social networks and Internet of things (IoT). He put forward the critical issues to upgrade big-data analysis, privacy and cloud security. He mentioned that motive is to achieve enhanced ubiquity, mobility, security, scalability and quality of service (QoS) of clouds and highly-visited social networks or datacenters. He evaluated the widespread use of clouds over massive datasets generated by e-business, social networks, sensors, RFID, GPS, etc.

His talk also revealed major R&D challenges and presented new approaches to preserving data privacy, assuring cloud security, and enhancing cyber intelligence. To remove the security and trust barriers in bare-metal or virtual clouds, he examined the top-10 security and privacy issues released by Cloud Security Alliance in 2012. Some new approaches and hidden opportunities are discussed towards the building of a trusted and intelligent cloud computing environment over both structured and unstructured big datasets. Finally, he compared the security and capability in BYOD (Bring Your Own Devices) solutions with those offered by the new BYOC (Bring Your Own Clouds) approach for inter-cloud (mash-up) applications.

Arvind ParthasarathiArvind Parthasarathi, President, YarcData delivered a talk on “Delivering on the Promise of Big Data”. He mentioned that the real promise of big data isn’t about merely doing analytics cost-effectively and at scale; it’s about discovery. Data discovery means uncovering hidden patterns from disparate sources without needing to know which questions to ask or the data relationships in advance.

He emphasized that the “Big Data” revolution puts a lot of pressure on IT budgets, which on average only grow by 5% per year. Still organizations have to cope with a 40% data/systems growth per year. He discussed the need to evolve analytics from tactical objectives toward strategic data discovery that heightens business performance, insights and protection by identifying previously undiscovered relationships and patterns in data. He also presented YarcData’s purpose-built system for data discovery offering a combination of hardware and software.

Rebecca D. CostaRebecca D. Costa, American Socio-biologist and Author talked on “The Big Deal about Big Data: a Socio-biological Perspective”. She argued that we adapt very slowly to change. We enter a high failure-rate environment (the number of wrong options is growing much faster than the number of right options). Complexity causes us to become bad pickers.

She said “In nature, any drive toward singularity is a drive toward extinction”. Efficiency is often a dangerous drive toward singularity because you start to remove redundancy. She suggested following to succeed in a high failing rate environment:
  • Diversity of solutions: venture capital model
  • Expect failure, “fail fast”
  • Leverage “Big Data” combined with mobile/interactive technologies
  • Crowd source (internal as well as external)
  • Avoid “belief-based” policy

Mario InchiosaMario Inchiosa, US Chief Scientist, Revolution Analytics delivered a talk on “Scaling R to Big Data Science”. He described current challenges and issues in big data analytics -- the talent gap caused by tool chains requiring myriad skills, lack of reusability across platforms, the cost and delay of writing advanced parallelized analytics from scratch. He also mentioned how Revolution Analytics is helping to address these challenges via Revolution R Enterprise (RRE).

RRE is a commercial distribution of open source R that includes Intel Math Kernel Libraries, select open source extension packages, Revolution’s ScaleR functions for scalable, cross-platform advanced analytics, and Revolution’s DeployR Web Services framework. They recently enhanced RRE to support in-Hadoop and in-Teradata analytics with big data implementations of advanced analytics and machine learning algorithms for exploratory data analysis, regression, classification, prediction, and unsupervised learning. In the talk, he also examined how RRE implements these technical advances, making big data analytics more powerful and accessible.

Highlights from day 3.