Big Data in Medicine - rather big noise?
It is true, that we are about to enter into new dimensions with Big data, but is also true, that many are mesmerized by its sheer magnitude and just follow the mass of euphoric, uncritical so called success stories. Here a good article which put much into a correct perspective - just a few examples:
- Big Data’s strength is in finding associations, not in showing whether these associations have meaning!!!
- Big Data analysis is largely based on convenient samples of people or information available on the Internet - so how about data quality?
- Big Data are observational in nature and are fraught with many biases such as selection, confounding variables, and lack of generalizability
So, dose this makes sense? Here you can read more
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4684636/
Chief Medical Officer & Co-Founder at we4 Impact
7 年Absolutely agree with Achim. I think big data offers lot of opportunities and it helps (insurance) experts to find new correlations and to ask new and interesting questions. But (and this is a but in big letters) correlations should not be confused with causality. And the larger the dataset, the more meaningless correlations will be found. And if data quality is not good , the more wrong correlations will be detected. Big data still needs real people to check for plausibility and potential biases.