Columbia University - Creating a Data Science Module That Fosters Upskilling and Professional Opportunities for Finance Students
Name: Perry Beaumont, PhD
Title: Lecturer, Columbia University
Country: United States
Organization: Columbia University, School of Professional Studies
Columbia University's School of Professional Studies is committed to providing diverse, innovative programs to enable students to succeed in their educational and professional development.
Excellence in Teaching
As a university lecturer, I encountered the challenge of sourcing real-world data in combination with enterprise platforms to help students bridge textbook learning with hands-on applications. At about the same time that I decided to simply create my own lesson plans in this regard, I was contacted by the university publisher Cognella to create resources inclusive of an online text that would provide precisely what I perceived to be a gap in learning resources.
As a result, I committed to working with Cognella to write a book (available in hard copy, and online) that would present case studies in data science, titled "Business Case Studies in Applied Data Science: Supply Chains, eCommerce, and Consumer Lending".
Columbia University uses the book for a graduate-level finance course (though oftentimes undergraduates will petition to be added), and the class is entitled Security Analysis. Security Analysis is the name of a book written by Benjamin Graham and David Dodd years ago on the topic of Value Investing, and Warren Buffet was a student of Benjamin Graham when Buffet was a student at Columbia Business School. Since the course is in finance and not data science, there is a mix of knowledge and comfort levels among students related to quantitative analysis, and that is an additional reason why the Dataiku platform is so consequential in the context of the Case Studies text; Dataiku's intuitive and easy-to-navigate platform makes it a joy for students to quickly grasp and apply a variety of analytical insights.
While the text version of the book is available for purchase now (at Amazon Books and elsewhere), the digital version will be used by universities globally for the first time this fall (and officially launching at Columbia University on September 6, 2022). A variety of marketing and promotional channels will be used by Cognella to assist with supporting distribution efforts.
With my previous exposure to the outstanding resources of Dataiku, I immediately began to write a module for the book that would involve the Dataiku platform. From a pedagogical standpoint, the visuals and intuitive foundations of Dataiku make it an amazing learning resource. With the invaluable assistance and guidance from @AdelaD, @DamienJ, and @CoreyS, a data science learning module was cobbled together using a Google e-commerce dataset and Dataiku's machine learning platform.
With brick-and-mortar retail, the average conversion rate is about 20%. For online e-commerce retail, the average conversion rate is closer to 2%. Accordingly, being able to improve an online conversion rate by as little as 0.5% translates into a 25% improvement in sales. By delving into this challenging data exploration, students have at times able to identify strategies for improving conversion rates by a magnitude of 0.25-0.60%, which is simply amazing.
By providing students with these types of opportunities to analyze data — the types of challenges that companies actually need to think about and solve — we are better preparing the next generation of professional leaders to be comfortable with using advanced tools and insights for better decision-making.
The impact from the learning modules extends not only to students' realizations of the ROI for analytics generally and particulars of a successful business challenge (i.e., improving conversion rates) but additionally, includes success measures related to upskilling and professional opportunities.
There has already been tremendous student feedback with the early drafts of the book who are relating that their experience with the case studies (and the module involving Dataiku in particular) helped them to stand out in a positive way during a job interview, as well as when being asked to participate on various employer teams created to examine analytic challenges.
The following was shared by a student with specific reference to the Dataiku module, and who was part of a 100-person cohort of initial reviewers of the new text:
"At first I was a little unsure of how I would do because my primary area of study is not super intensive into quantitative topics, and the approach used in Case Studies is really helpful with how it explains things at each step. Also the screenshots and explanations for how and why were great, along with the graphics (and especially with the recipes)."
Value Brought by Dataiku:
There is so much to say here regarding the value brought by Dataiku with this initiative.
First, there is the consideration that Dataiku offers a free one-year license to students, and this quite simply is amazing! And thank you Dataiku for making this investment in the future of those students who will soon be going into the professional world and benefiting from the exposure you have made possible for them with a better understanding of data science applications.
Second, there is the sense of a supportive community available to students with Dataiku, with the ability to pose questions and share knowledge in a collegial and welcoming way. Further, the ability to easily seek out answers also exists with Dataiku's expansive Knowledge Base.
Third, the learning resources provided by Dataiku, as with the Academy, offer tremendous online opportunities for students to go well beyond the introductory material provided in Case Studies to further explore their interests with instructional videos and the pursuit of online certifications.
Fourth, Dataiku performs an amazing service by helping students to see the myriad industries and applications of data science with its Stories portal, facilitating insights into the many ways students as future professionals can be a meaningful part of the data conversation with whatever career path they might ultimately choose.
Finally, the strong commitment and value that Dataiku attaches to Responsible AI really serve to provide a North Star for students with having a meaningful framework to see how knowledge and unique skillsets also embody a special responsibility that exists when models are accountable, architecture and infrastructure are sustainable, and data processes are governable.