Three Cornell alumni and affiliates were on-hand recently to share insights with current Big Red students about career life in data science. Hosted by the Cornell Statistics Graduate Society, the Career Panel for Data Science was held on Friday, May 4, and featured the following panelists:
• Cecelia Earls, (Cornell PhD, '14), currently a lecturer in Department of Biological Statistics and Computational Biology. She has prior work experience working as an actuarial consultant in a human resources firm.
• Kushagra Aniket (Cornell BA, '14), now working as a Research Analyst at Cornerstone Research.
• Matthias Kormaksson (Cornell PhD, '10), now working as a Senior Principal Statistical Consultant at Novartis.
What are required for data science jobs? Like degree, major, paper, internship, skills, etc.
The most important strategy is to search job postings to know their requirements. Generally, communication (for example, how to explain technical concepts/results to non-experts), collaboration, Python and machine-learning skills are important for many data science jobs. Specifically, in many industrial data science jobs, they are looking for individuals with at least a master's in data science related majors like mathematics, computer science, and statistics. You may be asked to show analytical skills by portfolio or competition. If you have shown enough skills but don't have a data science-related major, usually that’s also fine. To prepare for industrial jobs, internships and group projects that simulate a team working environment are recommended.
What are your criteria for choosing a career? For example, academia vs. industry?
The criteria include career interests, life situation and skill match (i.e., do I have the relevant skills for the position?).
What statistical methods do you use most frequently?
It depends on the field.
General advice from Prof. Earls:
(a) Learn basic statistical methods well, such as regression of all types.
(b) Become proficient in commonly used software for statistical analysis and data management. For example, SQL is very useful for big data.
(c) When doing a project, make clear your goal, estimation (e.g. does x2 have significant effect on y) VS prediction? Then, accordingly, choose the appropriate statistical methods.
In Kushagra’s role as a research analyst in industry: (1) data presentation and visualization are highly important. (2) experimental design: surveys. (3) standard statistical tools, including fixed effects regression, time series, limited dependent variable models, non-linear regression, forecasting, Monte Carlo simulation. (4) valuation tools to assess business impact of proposals, like DCF and comparables.
Could you share your moment of successful networking? How/where to start networking?
Try to connect with Cornell alumni, parents, friends etc. who work in your field of interest, since they are likely to offer help. Sometimes there are groups on campus that are well connected to alumni in a particular field. So, it may be helpful to join one of these groups (such as a club or fraternity in your field of interest). Linked-in can also help you find and contact alumni.
You can also attend on-campus career fairs. Drop your CV and you may get response from employers.
Panelists: Passion and skill-match matter.
Panelists mainly focused on matching our passion and skills with those of the job positions, which is essential.
Think about your long-term interests, what you are good at, and the type of life you like. Be brave to trial-and-error. You are likely to make several career changes before you find a good fit.
Communication and collaboration are highly important skills needed in nearly any job. Generally, useful skills include basic statistical methods, Python, machine learning, etc.
Search job postings to know whether your passion and skills match the job positions.
This event was hosted by Cornell Statistics Graduate Society and funded by Department of Statistical Science and Cornell Graduate Assembly (GPSAFC).