Deep Learning to classify big data

By Sahar Tavakoli

Our brains do an expert job of classification; it happens when we recognize people from their faces, categorize an object that we see, or predict the future state of an event. Proportional to  the complexity of an input pattern, the classification can be easy (for example recognizing the difference between a cat and a dog) or difficult, such as predicting the probability of two people becoming friends in a social network.

Human brain 2

The human brain’s deep architecture has been an inspiration for deep nets.

The human brain’s deep architecture, as an example of a successful agent in classification tasks, has inspired the development of deep nets, artificial neural networks that can learn tasks that contain more than one hidden layer between the input and output layers. This work has shown amazing result recently.

In deep neural networks, each layer consists of several nodes, called neurons. Edges in a neural network are between two consequent layers. Each edge has a weight and each node has a bias value. These neutral networks “learn” by applying training inputs to the input layer of the network and comparing them to the desired (and known) output. The input is weighted and modified until the desired outcome is achieved.

Neural networks have been around since 1970, but back then, they were successful only in simple pattern problems. Tackling more complicated patterns meant adding layers to the system and training with a large number of layers. This is what is now called a deep net, and it was not possible until recently, when computer scientists found better ways to train a deep net such as applying new types of activation functions to nodes.

Deep learning has shown amazing results—sometimes better than those of the human brain—in areas such as image processing, speech processing, social media analysis and biology. Since training a deep net is not possible with restricted amounts of data, having enough data is a prerequisite in using deep learning techniques. Social media, which creates massive amounts of data every day, is a good source of big data and therefore can help in further developing deep learning and neural networks.

During the 2017 International Conference on Social Computing, Behavioral-Cultural Modeling, & Prediction and Behavior Representation in Modeling and Simulation (SBP-BRiMS 2017),  I observed that several research groups are using this emerging method in the social computation field. These researches have different interests in using deep learning techniques, including the ability to predict links in social media, text analysis, predicting peoples’ feeling from pictures they share on social media, terrorism source prediction, news propagation modeling, and predicting peoples’ locations based on the information they share.

The conference made it clear that deep learning is a new and promising method in social computing research, but we must remember that working with big data and training a deep net requires the proper hardware, such as the most appropriate GPUs. These factors still restrict the use of deep learning, but the future is promising for this growing field.

20161015_205156Sahar Tavakoli is a PhD student in computer science at the University of Central Florida. She attended the 2017 International Conference on Social Computing, Behavioral-Cultural Modeling, & Prediction and Behavior Representation in Modeling and Simulation (SBP-BRiMS) with support from the South Big Data Hub. The conference was held July 5 – 8 at George Washington University in Washington, DC.

Visuals, storytelling help make sense of data


Panelists discuss data visualization at a recent workshop sponsored by the South and West Big Data Hubs.

By Mark Schroeder

Throughout human history, stories have helped us make sense of sequences of events in our lives, infer cause and effect relationships, and share them with others. Just as our own memories are fallible and retelling stories can shape how we remember events, data can be fallible too. Its value is shaped by the process used to collect it and can be incomplete, incorrect, or biased in some fashion. How can we use data to gain true insights about the world and share them despite these challenges?

Continue reading

Mobile Health Workshop sparks ideas for future research

by Wenbin Zhang

Wenbin ZhangAs a first-year PhD student in information systems, I have been working on mobile health (mHealth) related research since the start of my PhD program. The growth of mHealth has facilitated better and instantaneous health communication, which was not previously possible. The capabilities of mHealth platforms promise to enhance healthcare quality and assist people in achieving healthy lifestyles at reduced costs. Attending the mHealth workshop organized by the South Big Data Hub and the National Consortium for Data Science (NCDS), located at the Renaissance Computing Institute (RENCI) deepened my understanding of mHealth, simply by having the chance to listen to and participate in intense discussions with an interdisciplinary group of mHealth and technology experts. Continue reading

South Hub and partners to hold data-driven storytelling webcast this Friday

The American Association for Advancement of Science (AAAS) Science & Technology Policy Fellowship Big Data Affinity Group, in collaboration with the South Big Data Hub, West Big Data Hub, and The National Consortium for Data Science, are making this Friday’s data visualization and storytelling event available for virtual attendees. To learn more about the event, visit the website or read our earlier blog post announcing the event.

Data-Driven Storytelling: A Deep Dive into Visualization Techniques 
July 14 | 9:00 AM – Noon ET | WebCast
Join the Webcast:
Call-in number:1-415-655-0003
Event Number: 641 886 660 | Event password: dataviz​

Developing Standards for Mobile Health


Attendees at the mHealth conference discuss key issues, including mHealth standards, at a breakout session.

By Alex Cheng

I was honored to have the opportunity to attend the Mobile Health (mHealth) conference sponsored by the South Big Data Innovation Hub and the National Consortium for Data Science as a third-year graduate student in biomedical informatics at Vanderbilt University. My research focuses on using mHealth technology to improve the efficiency of outpatient clinic operations and the quality of care for patients.  Continue reading

Can wearable devices lead to better health outcomes?

Reflections on the South BD Hub mHealth Workshop

By Chenzhang Bao

mhealth-phone-e1498589419307.jpgIn recent years, mobile health (mHealth) has become one of the most popular health care movements for patients and providers. Consumers have embraced the use of mHealth applications in their daily lives through wearable devices, and use these apps to monitor their exercise routines, heartbeats, and sleep quality. The use of mHealth apps is critical for research into new mechanisms designed to improve the quality of patient engagement; a factor that has previously been hard to measure or even unobservable to providers. One important research question looks at the relationship between patients’ usage of mHealth devices, their engagement in their own health and the future health outcomes. Continue reading

The 2017 All Hands Meeting of the South Big Data Hub


When we launched the Big Data Innovation Hubs at the end of 2015, we could never have imagined that our mission of “breaking barriers, bridging solutions, and accelerating partnerships,” intense but rewarding work, would yield over 800 members—many of whom actively contribute to Hub communities of practice, dozens of productive partnerships, several funded new projects, and nearly 20 workshops. A year and a half later, on Friday, June 9, 2017, more than 75 people from across sectors and disciplines—academia, government, nonprofits, and industry—met at the Microsoft Chevy Chase Pavilion near Washington, DC, to assess the progress of the South Big Data Hub, and shape its future. It was a day of catching up on current efforts (some of which began at the first all-hands hub meeting), and sparking new collaborations.

Continue reading

Microsoft Research looks back at a year of successful collaboration with the Big Data Hubs


Vani Mandava of Microsoft Research (far right), with leaders of the Big Data Hubs, from left to right: Fen Zhao, NSF program coordinator; Lea Shanley, South Hub; Melissa Cragin, Midwest Hub; Rene Baston, Northeast Hub; Meredith Lee, West Hub; and Renata Rawlings-Goss, South Hub.

Microsoft Research understands that taking full advantage of big data and new data technologies requires more than developing new tools and technologies. To paraphrase Vani Mandava, director of data science for the research arm of the tech giant, it requires cross-disciplinary research that extends well beyond computer science, and collaboration among domain science experts, computing and data science specialists, and industry leaders in technology and other verticals.

Continue reading

Shifting the paradigm of care via mHealth

mHealth-3By Ashley C. Griffin

The South Big Data Innovation Hub and the National Consortium for Data Science (NCDS), in collaboration with the Institute for the Future and 10X Collective, held a workshop that brought together a diverse body of experts to identify and prioritize research challenges in data science and IoT cyberinfrastructure.

The workshop participants thoughtfully assessed a wide array of mobile health (mhealth) applications to address health disparities and their environmental influences within the research, legal, policy, environment, and clinical settings. Within the clinical setting, participants identified shifting the point of care to the patient using mHealth technologies as a key priority.  Continue reading