Data analysis and data mining have different thinking modes. Generally speaking, data analysis is based on objective data for continuous verification and hypothesis, while data mining has no hypothesis, but you should also give your judgment standard according to the output of the model.
When we often do analysis, data analysis needs to be more thoughtful and use a more structured and MECE thinking method, similar to IF else in the program.
But data mining is mostly large and comprehensive. The more data, the more accurate the model, the more variables, the clearer the relationship between data, and all variables are needed. Firstly, the variables in the sense of the model are selected (big and complete, many and precise), and then they are screened according to their correlation, substitution and importance, and finally all of them are thrown into the model. Finally, whether this method is reasonable or not is judged from the parameters of the model and the meaning of explanation.
The feeling of big data is not a large amount of data, nor is it complicated. These can be handled with tools and technologies, but it can judge rules in real time.
For example, the push of targeted advertising is big data, which can accurately push relevant information to you according to your previous browsing behavior, basically making you a database, not a data. But the data analysis we do is more about groups than individuals.
Therefore, the era of big data has also exposed various problems, such as data privacy, data killing, data islands and so on. This may be the reason why we now see that big data analysis pays more attention to technology and means.