Annual Forum

​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​Web Banner-01.pngWe had yet another wonderful Advanced Analytics and Predictive Modeling Forum this year, with nearly 150 attending guests. An enormous thank you to all of our speakers and sponsors for making this terrific event possible. We'd also like to thank all of our guests for providing us with valuable feedback that we can use to improve your experience at next year's forum. If you missed the forum and need your contact information updated or if you have any other questions, please email us at ihbi@cmich.edu. Thank you all for your continuing support of IHBI, and we look forward to seeing you again next year!



14th Annual Forum Presentations (2015)


John Elder, PhD
John Elder.jpeg
President and Founder of 
Elder Research
​​




"Top 10 Analytic Mistakes"

Deeply analyzing data to discover useful patterns has delivered enormous returns in many fields. But, it is easy to go too far and “torture the data until it confesses” and doom your findings to fail where they really matter: on new situations. A key to achieving quality is to avoid “worst practices”. Dr. John Elder will share (often humorous) real-world stories of common deadly mistakes. Be encouraged on how to achieve success through tales of barely averted disaster.

Download Presentation Button.jpg

Line.jpg  
Brian Hochrein
Coming Soon.png
Director of Applied Analytics at Truven 
Health Analytics
​​




"Statistical Graphical Deception"

As we continue to collect, analyze, and interpret more data using both classical and modern approaches, visualization has become a large part of the way we present results or obtain information. Since graphical construction can either help or hinder the audience’s understanding of the results, we need to be careful to not fall victim of the “dark side” by creating charts and graphs that are deceptive in presenting the truth. We will explore the visual decoding process, which is called “graphical perception”. We will look at what makes a graphical method successful or unsuccessful and potentially deceptive to the audience for which it is intended. The goal is to provide graphical techniques and strategies for both data analysis and data communication.


Download Presentation Button.jpg 

Line.jpg 
David Corliss, PhD
dcorliss_head_shot.jpg
Predictive Analytics Lead 
of IT and Analytics 
Services at Ford 
Motor Company
​​




"The Analytical Death of Personally Identifying Information (PII)"

Personally Identifying Information (PII) allows unique identification of an individual, associating a person with a set of data. As a result, PII requires a high level of security. However, today’s data mining and analytic tools enable unique identification of individuals on a large scale without the use of PII. A combination of data points, none of which identify a single person, can be combined to create a unique identifier. The ability to leverage analytics to create complex custom-built keys to identify individuals can make the concept of PII obsolete. The advent of UII - Uniquely Identifying Information - may lead to a reassessment of the security of personal data.


Download Presentation Button.jpg  

Line.jpg 
Doug LaLone, JD 
LaLone_Doug_color.jpg
Partner and Patent 
Attorney at 
Fishman, Stewart, & 
Yamaguchi PLLC
​​




"The Legal Side of Big Data"

The legal realm of Big Data and analytics is a new frontier that is quickly developing and is full of land mines. The private sector moves ahead at lightning speed engaging in sophisticated methods of data gathering and computer analysis in hopes of drawing out insights which can drive efficiency and progress.  However, industry is falling short of effectively managing privacy considerations, data ownership, model ownership, and Intellectual Property considerations. Worse yet, our legal system is even further behind industry by slowly setting policies and laws for governing the behavior of industry where consumer data is often given little consideration. In the era of Big Data hacks and leaks, by both US and foreign operatives, US companies are being subjected to much exposure. Doug LaLone will challenge us to consider the legal side of Big Data, where we are today, and the challenges facing businesses. 


Download Presentation Button.jpg 

Line.jpg 
Sadie Bell, MA
Coming Soon.png
Subject Matter Expert 
for Analytical Reporting 
and Dashboards at 
The Dow Chemical 
Company
​​




"The DOs and DON'Ts of Visualization"

Designing clean, clear, and compelling visual representations is a key component to sharing insight and enabling business decisions. Power BI is an end user tool designed by Microsoft to help data explorers and analysts quickly analyze and visualize data. However, tools of this nature always work better in a demo than in practice. Today, you will learn some of the DOs and DON’Ts of visualizations with a demonstration using a tool called Power BI and learn some key methods for detecting and preventing these mistakes from taking place or reoccurring. These principles can be applied across tool sets and are not limited to Power BI in any way.

 

[Download coming soon]​

Line.jpg 
Jo Porter, MPH
Coming Soon.png
Interim Director of the 
Institute for Health Policy 
and Practice at the 
University of New 
Hampshire
​​




"The Dark Side of Healthcare Analytics"

Over the past ten years, there has been a revolution to make health care policy and practice more data-driven. There is tremendous need for data to inform clinical practice, health policy, and consumerism in healthcare. All sectors of the industry seek better data, and the availability of data continues to improve. However, despite this expressed need for more data and, in many cases, the increased access to data, healthcare data are imperfect…to say the least. This session will focus on some of the common challenges in analytics for the healthcare industry, with particular focus on use of administrative claims data. The session will include discussion of challenges with analysis at the provider level, the adaptation of financial data to non-financial purposes, and the need to balance access with privacy. 


Download Presentation Button.jpg 

Line.jpg 
John Elder, PhD
John Elder.jpeg 
President and Founder of 
Elder Research
​​

"Target Shuffling (to Save Science)"

It’s always possible to get lucky (or unlucky). The central question in statistics is “How likely could this result have occurred by chance?” 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 over-fit 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 good news is an antidote exists! John Elder will explain the simple breakthrough solution that’s rarely employed, but is being rediscovered in leading fields. He will illustrate how to use the re-sampling method he calls “target shuffling” in multiple scenarios, showing how it calibrates results so they are reliable. Bottom line: Honest data science is needed to save experimental science!


Download Presentation Button.jpg 

Line.jpg 
Nolen Akerman, PhD​ 
Nolen.jpg
Senior Business Analyst 
of Global Analytics at 
Kellogg Company
​​




Speaker Panel: "Best Practices in the Workplace"

The forum speakers participate in a facilitated panel to discuss and answer audience questions regarding real-world application versus theory, and share their insights on opportunities and challenges for effective cross-industry utilization of Big Data.



Download Presentation Button.jpg

​​

​​​