Fourteen Expert Perspectives on Data Science
Last month, the National Convergence Technology Center (CTC) hosted a Special Interest Group (SIG) workshop at the HITEC conference entitled “Essential Skills for the Next Generation of Work: From Data Literacy to Data Analytics to Artificial Intelligence.” In this lively and interactive session – the final half hour was framed as a Jeopardy-style game show for attendees – data scientists Aaron Burciaga, Jordan Morrow, and Kirk Borne discussed data analytics workforce skills with a particular focus on the importance of students becoming data literate.
Data scientist Aaron Burciaga leads a faculty workshop on the essential data science and AI skills students need to know.
Below are some highlights of those presentations and discussions.
1. Students need to be learning foundational data skill and competencies, but too often program focus on teaching the point-and-click tools like Tableau. Students need to understand how to work with data to solve a specific problem.
2. Any information you can use is data. If it’s something that you can use to improve something, it’s data.
3. Employers need to learn to trust two-year colleges and stop their practice of only considering four-year graduates or advanced degree holders for data science jobs. While there is always a place for PhDs and Master’s degree when it comes to leadership and innovation, it’s the two-year graduates who are so valuable in doing the actual work.
4. China plans to be a global AI superpower by 2030. There is an urgent need in the U.S. to build that industry.
5. Consider the letter “T.” Students need a wide breadth of know-how across many skills and concepts (the horizontal bar of the T) but depth in a specific specialized area (the vertical line of the T).
6. Adding a certification – like the Certified Analytics Professional – to a degree is recommended. Think of the associate’s degree as the foundation and the certification as the specialty.
7. The CMMC (Cybersecurity Maturity Model Certification) is not only changing the cybersecurity landscape, but also encouraging “responsible AI” that is impartial, resilient, transparent, secure, governed, and accountable. Learn more about responsible AI here.
8. Business is not good at data science. Most do not know what to do with the data they have. There is no power in holding data. Business must take action on the data.
9. There is likewise a challenge in data scientists effectively communicating with the C-suite executives. For this reason, students also need to learn business, communication, and leadership skills. Those “soft” skills need to added to their technical skills.
10. There are four levels of data analytics. “Descriptive” describes the problem – as in, “you’re sick.” “Diagnostic” looks for trends and patterns – as in, “here’s why you’re sick.” “Predictive” offers possible fixes. “Prescriptive” provides an external solution. These steps have to be repeated over and over. It’s an iterative process with the goal of continuous improvement.
11. The number one factor for data analytic success is the company’s organizational structure. A holistic strategy is essential. Companies need to ask five questions: what am I trying to achieve, how do I get there, do I have the right tools, do I have the right skills, and do I have the right environment?
12. It’s important to strike the balance between the two extremes of making decisions purely on gut instinct without any data and making decisions blind to environmental context based purely on data. The human element still plays a role.
13. All employee levels need to be data literate: C-suite executive, mid-level managers, and entry level workers.
14. “Data skepticism” isn’t the same as “data cynicism.” It’s not about not trusting the data. It’s about examining with a skeptical eye how the data is used.
During this SIG workshop, Aaron discussed his vision of a “Blue Collar” AI workforce of farmers, miners, and welders trained by two-year technical colleges. To learn more about this concept, take a look at this 20-minute webinar.