Data Science or Statistics/Software Engineer
I am a M.S. statistics student about to graduate. I wanted to ask for advice about getting into a career after graduating. I am interested in data science and I see that there is a lot of opportunity. I have taken many classes and a few internships in the data science field as well. But, I am reluctant to get into the data science field. I feel like data science is a merger of 1/4 of the technical skills of computer science and 1/4 of the technical skills of a statistician and someone can even get around with knowing nothing about the math behind data science. I have learned a lot of technical knowledge in statistics and I feel I would be forgetting this if I chose a career in data science. Overall, I do not feel that data science requires a lot of technical skill and it is much easier for someone outside the field to get into than something like statistics or swe. Also, since everyone is getting to data science, I feel that a statistics career is a better option, but may come with lesser tc.
Stats by itself doesn’t have any career these days
There's working at a pharma company in a statistican/biostatistican role. Working in insurance as an actuary, healthcare as part of an institutional research team, ux researcher designing and analyzing experiments
Those are extremely boring in my opinion and use outdated, pedagogic methods over and over.
It’s the difference between 70k/yr vs 300k/yr. Choose wisely!
Thanks for everyone sharing their knowledge/experience.
> Overall, I do not feel that data science requires a lot of technical skill and it is much easier for someone outside the field to get into than something like statistics or swe You’re confusing DS for Data analytics. Applied Scientist/ML engineer has the same bar coding wise as SWE in most places, and Research scientist has a high bar publications wise.
Correct me if I am wrong, but for me using Keras to build models is not a lot of lines of code. I only have academic experience in building deep learning models, but I have understood/written code for CNN, RNN, LSTM, GRU models even in Pytorch and they don't look very complicated coding-wise. Now deep learning in academic research in published journals that coding is extensive, but I think this group of DS people represents a small portion of DS people. This is maybe the kind of coding skill you meant by "Applied Scientist/ML engineer". I think this is not the majority.
Api != product. Someone still has to build the product. You will override defaults, extend classes, implement papers. All the skills of SWE + ml knowledge. That’s a DS job worth having