3 ms·
How is data science not just statistics, rebranded? When I studied statistics in university we studied clustering, regression, hypothesis testing, confidence i
by jordigh 3y ago
How is data science not just statistics, rebranded?
When I studied statistics in university we studied clustering, regression, hypothesis testing, confidence intervals, visualisations, charting, general linear models... how is this different from data science?
Is it just because we have more computing power, so we rebranded it? Instead of doing frequentist analyses because those are easy to do with smaller computers, now we do more Bayesian stats, on bigger computers, and therefore call it data science?
- topkai22 3y agoFrom the one data science course I took: compared to a traditional stats class, there is a bit more focus on data acquisition and pre processing, and more usage of our of the box tools, a bit less on some of theory (I don’t recall hearing about type I and type II errors). Of course, since it was a professional course there is probably the assumption that students already know these things…
- slt2021 3y agostatistics assumes you have data prepared and ready to work on. data science is more practical in a sense, you learn more to prepare real-life data for analysis. also statistics coursework doesn't cover ML part (at least the stats course that I took)
- tnecniv 3y agoI’ve periodically looked into it the differences and mostly it seems to come down to priorities. Statistics prioritizes mathematical rigor all the way through the process from collection to analysis. ML / Data science has more algorithmic concerns. Even the theoretical aspects of the two fields emphasize that distinction.