Day 1: Monday, September 15, 2014 |
REGISTRATION & BREAKFAST
8:15-8:30am
CONFERENCE CHAIR WELCOME REMARKS
[ Top of this page ] [ Agenda overview ]
8:30-9:30am
KEYNOTE
The Peril of Vast Search (and How Target Shuffling Can Save Science)
It's always possible to get lucky (or unlucky). When you mine data and find something, is it real, or chance? The central question in statistics is "How likely could this result have occurred by chance?" Ancient geniuses devised formulas to answer this question for multiple special-case scenarios. Yet, their calculus only applies to quaint, handmade analyses, where only a few hypotheses are considered. However, modern predictive analytic algorithms are hypothesis-generating machines, capable of testing millions of "ideas". The best result stumbled upon in its vast search has a much greater chance of being spurious. Such overfit is particularly dangerous, as it leads one to rely on a model molded to the data noise as well as signal, which usually is worse on new data than no model at all. The problem is so widespread that it is the chief reason for a crisis in experimental science, where most journal results have been discovered to resist replication; that is, to be wrong!
The good news is an antidote exists! Dr. Elder will explain the simple breakthrough solution -- still rarely employed, though newly being re-discovered in leading fields. John will illustrate how to use the resampling method he calls "Target Shuffling" in multiple learning scenarios, from model fitting to data exploration, showing how it calibrates results so they are reliable - essentially providing an honest "placebo effect" against which to test a new treatment (finding).
Bottom line: Honest Data Science can save Experimental Science!
9:30-10:00am
Sponsored Session
Anticipatory Analytics: The Path Towards Future Truth
We live in a data-driven world where today's government agencies are tasked with seizing opportunities from big data and analytics to be more efficient with fewer resources and budget. The massive amounts of unstructured and structured data available today allows agencies to solve unique business challenges with even greater speed and precision.
D&B's Chief Data & Analytics Officer, Paul Ballew will discuss how advanced analytics is changing how government operates today. Use of cutting-edge anticipatory analytics by some agencies are providing new levels of transparency and foresight critical to their missions. This path toward future truth – and actionable insight empowers agencies to make smart, data-driven decisions – and ultimately delivers tangible benefits to the American taxpayer.
BREAK
Track 1 Sponsored By 
Track 2 Sponsored By
10:35-11:20am
Track 1
Finding Patterns of Emergence
Can scientific journals today tell us what technology will look like in the next five years? Can crowdsourcing methods better estimate when S&T milestones will be reached? The FUSE (Foresight and Understanding from Scientific Exposition) and ForeST (Forecasting Science and Technology) programs are trying to find out---by predicting technical emergence using information found in published scientific and patent literature and from aggregated human judgements. Join Dr. Dewey Murdick as he discusses the goals and insights of these programs, which are led by the Intelligence Advanced Research Projects Activity (IARPA), the high-risk, high-reward arm of the Office of the Director of National Intelligence.
10:35-11:35am
Track 2
Developing an Analytical Framework & Measuring Risk at the SBA
The use of predictive analytics has modernized the SBA and brought the agency to the forefront of the government lending community. This session will detail the past, present and future of SBA's credit risk management and small business lending policy – and the impact predictive analytics has on fulfilling SBA's mission of providing assistance to the small business community. This progress is in the development of the Lending Risk Rating, which modernized SBA's lender ranking system – and, more recently, the SMART and PARRiS protocols. Advanced scoring and analytics have also enhanced the agency's oversight capabilities.
[ Top of this page ] [ Agenda overview ]
11:20am-12:05pm
Track 1
From Analytics to Action and Results – A Law Enforcement Perspective
Law enforcement agencies play an important role in eliminating fraud, waste, and abuse. Data analytics is a critical tool in these efforts. Learn how HHS OIG is leveraging data analytics in its fight against health care fraud and turning analytics into action and results.
Gary Cantrell Deputy Inspector General for Investigations, Department of Health and Human Services
Office of Investigations (OI)
11:20am-12:05pm
Track 2
Deep-Dive Geospatial Case Study from Data Science DC
The data science revolution is caused by a fundamental shift in the cost of building analytical systems, and the complexity of learning to use them. Now, a single person can quickly perform data acquisition, integration, prediction, and presentation tasks, over a weekend that previously would have taken teams of people months. To demonstrate this, I will walk through how I quickly got a good-enough answer to a question for the Data Science DC Meetup, a volunteer organization that holds monthly professional education and networking events. Where, on a map, should I schedule events, to make most people at least sometimes happy with the location? This location analysis question is similar to problems faced repeatedly by companies and other organizations. Where should I put my regional offices, or chain of convenience stores, or satellite dishes? Techniques that will be discussed, complete with examples of R code, include: working with latitude/longitude data, constructing geometric cost functions, mapping, global function optimization, and dynamic report generation.
[ Top of this page ] [ Agenda overview ]
12:05-1:00pm
Lunch & Learn
Text Analytics Applications for Government
Sponsored By: 
Whether it's a one-time initiative for a special project or a solution for day-to-day decision making processes, unstructured data has great potential for government agencies. In this session, we will explore how federal, state and local government agencies use Text Analytics to unlock unseen insights in vast repositories of digital documents. Case studies and step-by-step demonstrations will illustrate how Text Analytics can help you:
- Address counterterrorism, health analytics, tip-line prioritization, fraud detection, audit, research and grant analytics, insider threat and more
- Detect topics, trends and patterns in freeform text
- Enable efficient insertion of subject matter expertise to optimize the text analytics process
- Enhance or enable predictive models with text data
- Visualize your text analytics results and make decisions through ad-hoc data exploration and interactive reporting
12:05-1:00pm
Lunch
Sponsored By: 
1:00-1:45pm
KEYNOTE
Predicting Predictive Analytics Savings and The Future
The Fraud Prevention System has provided valuable savings for the Medicare program; progress continues as CMS harnesses the FPS for more purposes and measures the results of its use quantitatively. These results, certified by the OIG, are reported in the most recent Report to Congress, issued in June of this year. Ms. Handley's discussion will cover the savings methodology, its promise and pitfalls, and the describe how the Fraud Prevention System is growing differently in the future into a multi-functional program. As CMS continues to refine its predictive analytic approach beyond a system that takes action against fraudulent providers, CMS expands FPS' potential and ultimately its yield as a multi-faceted tool.
1:45-2:45pm
PLENARY SESSION
Abstract to Concrete: Demystifying the Model Building Process
Predictive models are often discussed in an almost magical sense. In reality, models end up being countless lines of boring mathematical equations that a computer can understand and use to quickly calculate a model result. Model results can vary from numerical to a classification to a probability to inferences about something. This session walks through the process of marrying a business problem with some data, creating predictive models, and producing a result. The intended audience are those who have wrestled with forming a concrete understanding of predictive models from more general and abstract concepts.
[ Top of this page ] [ Agenda overview ]
2:45-3:00pm
Sponsored Session
3:00-3:45pm
BREAK
3:45-4:30pm
Track 1
Finding a Lost Person and Evaluating State Budget - Case Studies in Government Analytics
The Institute for Advanced Analytics created the nation's first Master of Science degree in Analytics. One major portion of the degree is an industry or government sponsored practicum project. Over 50 different projects have been solved over the past seven years. This presentation goes over two case studies from government agencies. The first case study tries to find a lost person in a city where a severe illness has broken out and they are thought to be the cause. The second case study helps a state agency plan their budget for the future.
Jason MillerFederal News Radio
3:45-4:30pm
Track 2
Geospatial Predictive Analytics Case Studies: Northern VA Sniper and LRA Leader Joseph Kony
Geospatial predictive analytics leverages social science research, commercial best practices, and advanced analytics to guide information-based decisions in the operational environment; effectively moving analysis from counting and reporting, to solution-focused collaboration and insight. This approach emphasizes empirical models and behavioral analysis, with a focus on time, space, and the nature of the incident or threat in support of targeted allocation and optimization of scarce resources. The ability to proactively identify and characterize actionable trends and patterns represents an important paradigm shift in operational security analytics. Pattern analysis of the Northern Virigina Military Shooting Series, and the Lord's Resistance Army (LRA) will be presented as operational use cases.
_McCue.jpg)
[ Top of this page ] [ Agenda overview ]
4:30-5:15pm
Track 1
Case Study: Chicago Department of Public Health
Transforming Food Inspections Using Predictive Analytics
The Chicago Department of Public Health is innovating to lessen the burden on taxpayers and improve public health by engaging predictive analytics in their food safety protection program. Using the frame of analytics, the department partnered with development and coding organizations to build a database. Data (including building environments, 311 calls, previous sanitation code violations, social media, previous building code violations, sourcing of food, lack of alley lights, population density, near-by construction, complaint history of other establishments owned by the same person/group, and many other variables) are fed into a predictive model that outputs risk scores for each establishment allowing targeting of inspectors
4:30-5:15pm
Track 2
How Luxury Cars and Predictive Analytics Impact & Protect the Integrity of Your Public Assistance Programs
The New York City Human Resources Administration (NYC HRA) Investigation, Revenue and Enforcement Administration (IREA) wanted to more effectively triage fraud within its Medicaid program. To that end, it designed a study that analyzed certain high-risk indicators, such as asset ownership, to see if they could improve investigative outcomes by increasing efficiencies and better focusing resources. The study found that the indicators were indeed predictive, and their usage could increase the probability of its investigative success by 71 percent, reduce investigation time by 21 percent and reduce costly and time-consuming interviews by 56 percent.
Saratu GharteyChief Program Accountability Officer, NYC Human Resources Administration Investigation
Revenue and Enforcement Administration
5:15-6:15pm
NETWORKING RECEPTION
[ Top of this page ] [ Agenda overview ]
Day 2: Tuesday, September 16, 2014 |
7:30-8:15am
REGISTRATION & BREAKFAST
WELCOME REMARKS
8:20-9:05am
KEYNOTE
The Health Data Initiative: Past, Present, and Future
In 2010 the Department of Health and Human Services launched its efforts to make the vast array of data resources it curates openly available for public consumption. Since then, the initial tranche of 30 data sets has grown to 1600 data assets, and counting. The Departmental efforts to release data for the purpose of sparking innovations in healthcare and the delivery of human services is known as the Health Data Initiative (HDI). The mission of the HDI is to help improve health, healthcare, and the delivery of human services by harnessing the power of data and fostering a culture of innovative uses of data in public and private sector institutions, communities, research groups and policy making arenas. Damon Davis will discuss highlights of what the HDI has accomplished since its inception, including successful applications by private industry, the program's current status, as well as what the future holds for the HDI.
[ Top of this page ] [ Agenda overview ]
9:05-9:50am
Practitioner's Panel
Applying Analytics in the Regulatory Environment
Session description is coming soon!
9:50-10:20am
Sponsored Session
Acceptable Big Data Analytics: Required Features for Successful Deployments
Perhaps one of the most widely discussed points of conflict and discussion regarding advanced analytics is privacy, and the right to control personal information. While the general debate of what policies are most acceptable, effective, and defensible is covered widely in various domains as well as the popular press, it is clear that to embed advanced analytics to optimize any processes requires security, model lifecycle management, approval processes, audit logs, and general transparency regarding how and why analytics lead to specific actionable outcomes. This presentation will review the features that must be considered in practically any application of advanced data mining or predictive modeling solutions to be successful – in the sense of being acceptable and compatible with the policies that govern the use of these methods when applied to individuals or groups. The discussion will also highlight the special challenges associated with current and emerging big-data storage and analytics, cloud storage and data integration, and other data technologies.
Sanjeev KumarDirector of Analytics and Business Intelligence Solutions Innovation
Dell Software Group
10:20-10:50am
BREAK
Track 1 Sponsored By
Track 2 Sponsored By
10:50-11:35am
Track 1
Case Study: Regulatory Oversight Using Predictive Risk Models at the British Columbia Securities Commission
The British Columbia Securities Commission worked with an analytics consultant to design and build four Risk Models that prioritize regulatory reviews for major processes, using predictive math. Come hear how a small regulatory organization overcame major hurdles with data and custom user interfaces to implement these models and adopted new ways of overseeing securities market participants.
Christina WolfChief Economist and Director of Economic Analysis
British Columbia Securities Commission
10:50-11:35am
Track 2
Building a Data-centric Strategy & Roadmap
As businesses struggle with the data flood, it is even more critical to focus on data as an asset that directly supports business imperatives. Organizations across most industries attempt to address data opportunities and data challenges to enhance business unit performance. Unfortunately, the results of these efforts fall far below expectations due to haphazard approaches. Poor organizational data management capabilities are the root cause of many of these failures. This presentation covers three lessons, which will help you establish realistic plans and expectations, and help demonstrate the value of such actions to internal and external decision makers.
[ Top of this page ] [ Agenda overview ]
11:35am-12:20pm
Track 1
Proactive Risk Monitoring Using Advanced Analytics at Ginnie Mae
Government National Mortgage Association (Ginnie Mae) is the guarantor of $1.5 trillion outstanding securities backed by over 8 million loans. Currently there are over 300 active Issuers/servicers in Ginnie Mae's program. Failures of these financial institutions poses an extensive risk and potential loss to Ginnie Mae and U.S. taxpayers. In recent years, Ginnie Mae has been building and expanding its capabilities in proactive monitoring of the risks, such as portfolio risk, credit risk and counterparty risk, by leveraging data sources from both inside and outside of the organization, developing sophisticated risk models, and enhancing enterprise-wide access to business intelligence reporting.
11:35am-12:20pm
Track 2
Predictive Policing: Preventing Crime
with Data and Analytics
Police departments, like many other government entities, are increasingly relying on data and analytics to operate more effectively and efficiently. In this presentation, Dr. Jennifer Bachner, Director of the MS in Government Analytics program at Johns Hopkins University, discusses how officers and crime analysts are using predictive analytics to deter criminal activity and solve open cases. In addition to discussing predictive techniques such as clustering and journey-to-crime analysis, she will present case studies of the experiences of Santa Cruz, CA; Baltimore County, MD; and Richmond, VA.
While this presentation focuses on the use of prediction for preventing crime in local communities, these techniques and tools can also be applied to other policy arenas, such as the efforts by the Department of Housing and Urban Development to predict and prevent homelessness and the Federal Emergency Management Agency's efforts to identify and mitigate communities vulnerable to natural disasters.
Dr. [ Top of this page ] [ Agenda overview ]
12:20-1:15pm
LUNCH
12:20-1:15pm
LUNCH & LEARN
Optimizing Decisions with Information Builders and Predictive Analytics
Traditional reporting provides a clear picture of the past, but has little power to shed light on the future. Combined with the power and flexibility of the Information Builders' Business Intelligence platform, predictive analytics allows customers to bridge the gap between backward- and forward facing views of business operations. In this session, we will dive into specific federal and commercial case studies with predictive analytics, and highlight how we have helped achieve quantifiable success, reduce cost and simplify maintenance.
1:15-2:15pm
KEYNOTE
The Need for Analytic Innovation and Managing Barriers
Dean Silverman, Senior Advisor to the Commissioner and Director of the IRS's Office of Compliance Analytics, will discuss how the IRS created a small team to help lead big changes - stopping billions of dollars in improper payments and refund fraud, strategically reducing a hundred billion dollar element of the Tax Gap, and improving core compliance processes. He'll explain why "big data" is only useful to an agency if its insights can be digested by front line operating divisions, and how building a "test, learn and adapt" culture creates capacity for change and innovation. Finally, he'll share how the lessons learned by the IRS can apply to your agency, helping to create a "learning government" to better serve the American public. Join him as he discusses The Need for Analytic Innovation and Managing Barriers with Federal News Radio's Jason Miller.
Jason MillerFederal News Radio
[ Top of this page ] [ Agenda overview ]
2:15-3:00pm
Track 1
Transforming Government Oversight: Collaborative Analytics
Early Returns on Investment in analytics projects combined with the passing of the Data Act means the time is right to invest in analytics. Now that we've experienced some valuable returns we see the next logical step is to help to move analytics to an open platform where others can benefit and contribute. We believe government oversight can be transformed by collaborating with others in the oversight community.
2:15-3:00pm
Track 1
Enabling IT Risk Management with Analytics
Fannie Mae's Operations and Technology Risk Management Office is in the process of maturing analytic capabilities to enhance the proactive risk management program. The program is tasked with helping the organization solve the challenge of improving the safety and soundness of the IT environment while empowering a high volume of change (i.e., agility). The presentation will focus on how we have worked to meet that challenge by driving behavior and culture change within a risk adverse organization.
Jason MillerFederal News Radio
[ Top of this page ] [ Agenda overview ]
3:00-3:30pm
BREAK
3:30-4:15pm
Track 1
Big Data for Small(er) Campaigns
President Obama's 2012 reelection campaign was the first ever to widely use statistical modeling and experiments to drive critical decisions. Since that election, smaller campaigns have aspired to use these techniques in similar ways, but lack the resources that Presidential campaigns have to build their own in-house analytics departments. Daniel Porter and Harrison Kreisberg of BlueLabs, both veterans of the Obama campaign's analytics team, will discuss some of the innovations in voter targeting that they developed on the Obama campaign, and their work at BlueLabs refining these techniques to make them available to non-Presidential campaigns--most notably Terry McAuliffe's successful 2013 campaign for governor of Virginia.
3:30-4:15pm
Track 2
Delivering Anticipatory Analytics at NGA
The National Geospatial-Intelligence Agency (NGA) delivers world-class geospatial intelligence that provides a decisive advantage to policymakers, warfighters, intelligence professionals and first responders. Anyone who sails a U.S. ship, flies a U.S. aircraft, makes national policy decisions, fights wars, locates targets, responds to natural disasters, or even navigates with a cell phone relies on NGA.
A key strategic objective for the agency is, by 2017, to provide an innovative range of anticipatory analytic products and services that make a measurable difference in outcomes for customers. Key to achieving this goal is the agency's use of the IBM SPSS product suite. In this session, learn how NGA is using this suite to tap legacy databases, build advanced analytics scenario's against both structured and unstructured text, and deliver embedded insights to its customers.
4:15-5:00pm
The Role of the Interpreter: Standing between Vision Leaders and Action Technicians are the Interpreters - Critical for Successful Analytics Deployments
What are the stumbling blocks to successful deployments? Why do deployments fail when the scientists walk away? Too often, the vision is established and the technicians hired to act, but they struggle to understand each other once the job is done. Critical to long-term success are the interpreters who understand the business of the organization but also have the technical prowess to facilitate the scientists. Deployments must seek the interpreters, and organizations must seek long-term value propositions to retain these critical people.
4:15-5:00pm
Track 2
Utah HR's Journey into Predictive Analysis
The State of Utah's Department of Human Resource Management is attempting to predict certain facets of its workforce including turnover, leave usage and manager success. As these steps are being introduced, those leading the change have faced some resistance and skepticism. (I know, in government??)
Retrace our path with us as we highlight the obstacles and successes we've encountered and then finish with a vision of where we are going as we look to the future.
6:30-9:30pm
Big Data Roundup
Big Data Roundup is not a traditional forum or mixer. We focus on "rounding up" local data centric folks to talk about related Big Data technologies that help manage, integrate, and analyze data. Following some short talks, attendees are encouraged to share thoughts in open discussions with community-proposed topics, including Lightning Talks & Roundup Sessions.
Round 'em up! Data engineers, database developers, enterprise architects, developers, analysts, data mining and business intelligence professionals are encouraged to attend and mix with other members of the community. We are actively seeking people who want to give a Lightning Round presentation that provides some perspective.
To find out More Information and to Register click here.
[ Top of this page ] [ Agenda overview ]
