Firstly, all samples rather than sampling analysis. With the development of technology, we are able to process massive data, from which we can get more reliable result than through sampling analysis. So we can give up sampling analysis in most cases.
Secondly, efficiency rather than accuracy. At the age of big data, we concern more on efficiency rather than accuracy. In other words, we should allow faults to improve efficiency. Comparing with massive data, some faults do not influence the final result.
Thirdly, correlation is as important as causality . We can’t deny the importance of causality ,but sometimes it is difficult or unnecessary to explore it. For example, the recommendation system of Amazon doesn’t know why the customer who likes Hemingway’s works is likely to buy Fitzgerald’s works, it just recommends and helps Amazon sell more than 100 times books than before Amazon using it. From this example, we can conclude that correlation plays an important role at the age of big data, so please don’t overlook it.
Finally, let’s review the three peculiarities of big data. …………Please remember them, maybe they will help you one day.
That’s all, thank you.
