GAB Forum: Machine Learning for Risk-Based Enforcement

This month's Government Analytics Breakfast Forum welcomed Jennifer Diamantis, Associate Director and Chief, Office of Market Intelligence and Austin Gerig, Assistant Director, Office of Research and Data Services from the Securities and Exchange Commission (SEC) to present their talk, Using Data Analysis and Machine Learning for Risk-Based Enforcement.

The presenters discussed how the SEC is using analytics to detect and gather evidence of insider trading and other market abuses.  Analytics allow the SEC to identify suspicious activity without receiving specific tips or complaints.  Further, the SEC is now better able to identify other individuals who may be involved with known suspicious activity. 

The SEC relies on a variety of statistical methods to make sense of the data they continuosly receive from the entities they regulate.  Using mostly open-source software such as R and Python, the SEC uses machine learning, text analytics, cluster analysis and network analysis to detect anomalies and abnormal behavior that may warrant further investigation.

The speakers were careful to note that these methods may, on occasion, produce false negative or false positives, so the results are carefully interpreted by those who have significant experience in the financial industry. 

The increased reliance on analytics has helped the SEC save resources by empowering them to accomplish more with less human capital.

A full recording of the presentation can be viewed here


Dr. Jennifer Bachner Discusses Best State Capitals on WalletHub

Director of the MS in Government Analytics program Dr. Jennifer Bachner discusses how to evaluate the nation's state capitals in WalletHub's 2018's Best State Capitals to Live in.  For example, in response to the question of whether residents of capital cities are more likely to be politically engaged, Bachner explains:

Given the work of political staff, advocacy groups, lobbyists and other professionals in state capitals, you can expect to find a higher rate of engaged residents in a capital city.  There are also reasons to suspect that even those not directly employed in the political arena may be more politically engaged.  For example, political science teaches us that higher levels of social capital are linked to more political engagement, and many state capitals boast tight-knit communities.  Further, research shows that political knowledge leads to an increase in political engagement, and residents of state capitals are more likely to possess higher levels of political knowledge as they work with or even interact socially with other residents whose professional lives revolve around state government affairs.

Read the full article here.


MA in Government Faculty Al From featured in Washington Post Magazine

Al From Al From who is co-teaching a class with Alice McKeon on Political Ideas, Strategy and Policy Implementation was featured this past weekend in The Washington Post Magazine -- see article.  He along with other political strategists discuss how the Democratic Party could benefit from the infusion of more moderate voices just as Bill Clinton did in the 1990s.  


GAB Forum: Melvin Greer, Chief Data Scientist at Intel

Today's Government Analytics Breakfast Forum welcomed Melvin Greer, Chief Data Scientist at Intel Corporation for his talk, Artificial Intelligence: Driving Innovation in Government.

Mr. Greer spoke about the many ethical and legal implications of using big data and automation in government.  For example, data collected on children, such as their academic performance and indiscretions, might significantly influence their prospects in the workforce later in life.  Is this a fair and appropriate use of data?
Mr. Greer also spoke about the need for high quality data science programs in higher education, and emphasized that these programs should focus on meaningful applications of data science methods to challenges across government.  He discussed some of the very real problems related to water, transportation, cyber security and healthcare that can be addressed with emerging analytical approaches.
It's useful to distinguish between different areas of data science, namely analytics, machine learning and artificial intelligence.  Each area is best suited to address different types of problems.  Researchers must devote serious consideration to whether a particular approach is useful for the problem at hand before diving into the analysis.
Many thanks to Mr. G1reer for a fantastic presentation.  A video recording will be posted here shortly.

GA Alum Publishes Book: The Art of Being Artificially Intelligent

One of our distinguished Government Analytics alumns, Zachary Hanson, has just published a new book: The Art of Being Artificially Intelligent: A Millenial's Guide to Faking It Till You Make It.  About the book: 

It can be difficult to attain lofty goals and aspirations, especially for the millennial generation. The Art of Being Artificially Intelligent: A Millennial’s Guide to Faking It Till You Make It provides an outline of six traits that can help millennials—and individuals of other generations—advance their careers and reach those aforementioned goals and aspirations. 

Through learning the importance of self-reliance, mental cadence, grit, mentorship, knowing your value, ongoing education, and terminal entrepreneurship, you’ll find the tools needed to break the stigma of being an “entitled” millennial and build a foundation upon which you can reach even the loftiest of goals.

You can purchase the book here.

Congrats, Zach!