applies, with special points of emphasis for data science positions.
Research Experience
Publications: Consider having a separate section for research experience/publications. For each research project, describe the data source (with an emphasis on the size of the data source), the methods you used, and most importantly the impact of your work (key metrics, key implications of results). Try to add as many links to ArXiv’s (both published and unpublished)
Titles: Try to mention the professor you are working under and also any relevant scholarships (e.g. your role can be presidential research scholar in machine learning under Professor ______ in the ___ lab).
Personal Projects/Experience
As mentioned above, data scientists must be quite versatile. For any area of DS that you have not covered, you should try to do personal projects/other internships to cover these areas. For any DS project, you should generally show:
Data Understanding: Show that you know your dataset very well. Make visualizations on the data and analyze their distributions. Discuss any key takeaways you may have. Dive into what data you have and what new features you can make (e.g. a new feature that takes salary and divides by the number of people in a household may tell you the quality of life of a family).
Note: Many people make the mistake of trying to slap random ML/stats methods onto whatever dataset they are given. Do not do this.
Technical Rigor: Show that you know your data science methods very well. Describe the processes by which you select certain ML/stats methods over others. Have solid frameworks for tuning parameters of your models (don’t just guess random one’s). List your assumptions for statistical analyses. Build custom deep learning architectures that make sense for your data problem. Show advanced understanding of metrics by choosing one’s that make sense for your data problem and making custom ones if no existing metrics make sense (e.g. trying to prevent false negatives in cancer detection). For all of these, justify your decisions to show that you aren’t lucky that these methods are working but rather, all your thoughtful decision-making has led to good predictive models.
Bottom Line Impact: Show that your analysis and predictive methods actually mean something. Businesses don’t hire data scientists to play with their data. Businesses hire data scientists so that they can derive insights that drive decisions and build models that can enhance their existing products. For every analysis you do and predictive model you make, justify why it’s good (does it do well on a specific metric?) and how this helps the company.
Want to print your doc? This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
Resume