Each newsletter a faculty member shares a relevant article, dashboard, visualization or video. This newsletter's faculty pick is Dr. Wende Mix, associate professor in the Geography and Planning department.
Why I chose this article and why should others read it?
ESRI, undoubtedly the largest global company for GIS, recently has promoted spatial analysis using AI and ML. For those of us who have been ‘doing’ GIS/spatial analysis for decades, many of the applications appear on the surface to be examples of the types of analysis that we have been doing all along. This perspective often is an issue for data science, in general.
The interviewer asks, specifically, why the examples being given are, in fact, different! In the videos associated with the interview, Wendy Keyes gives some very interesting and thought-provoking examples of how AI enhances basic spatial modeling approaches. Her argument includes the use of more and different data in an analysis and, an idea I found very interesting, changing the perspective of the analysis to the individual as opposed to the company. Human behavior is very complicated to model, so using AI, ML, or deep learning methods along with spatial analytics offers exciting possibilities.
Way back (when dinosaurs roamed the earth!) my dissertation was related to route choice modeling to predict driver’s destinations in the event of non-recurring incidents requiring detours. In short, dynamic route choice models. Forty years later, I can see the potential of GeoAI to better address this issue thus improving efficiency and sustainability of our transportation networks in major urban areas.
Are you one of the over 300,000 people who play Wordle daily? Then check out this article where Robert Lesser scraped Twitter to find out whether Wordle is really getting harder and what words have people most stumped. These visualizations show the popularity overtime, which words were the hardest to guess, and on average how many tries it takes to get the right word.
Nicole is aSenior Business Analytics & Reporting Analyst for M&T Bank where she provides analytical and modeling support that provides insights into managing the Consumer Deposit portfolio.
How did the PACM program contribute to your success: The exposure to programming with SAS, SQL and python languages and using those languages to complete various modeling projects.
What was your favorite part of program?:Creating life-long friendships with my classmates!
What advice do you have for current students?: Take advantage of as many opportunities as possible that will provide you with additional opportunities to learn a new application (tableau, Power BI, etc.), programming language or modeling experience.
DSA and PACM alumni, if you are interested in sharing information for a profile please complete this survey.
"A.I. is only as good as the data it is fed." - Mr. Monteith
Image recognition software can be bias if it is trained with bias data. This team is working to counteract this by collecting data from those who are familiar with the cultural background of these images.
Data Literacy for All
Check out this article from the Harvard Business Review by Thomas Redman, which talks about the importance of regular people in data related work. Including people outside the data analytics team is important in each step of the data life cycle.
This is article suggests ways of incorporating data ambassadors in each department to help improve your data projects.
Buffalo State Data Talk
Check out the most recent episode of Buffalo State Data Talk. Episode 18: Data Governance in Higher Education with George Firican, Director of Data Governance and BI at the University of British Columbia.
In this episode George Firican, Director of Data Governance and BI at the University of British Columbia and founder of LightsOnData talks about how his team uses data to help fundraise for the university. Listen to the episode to learn some of George’s favorite resources for budding data scientists and tips for growing your career.