This chapter is discussed about Machine Learning vs Statistics as data analysis to solve the prediction of the data, the correlation of the data, or making a good hypothesis
-
We can use statistic with significance test
-
Say in all 1000 people in our office, extract 10 people.
-
This could be office's favorit color, or maybe we just coincidence to pick 10 people who all pick "Blue"
-
Suppose we have different web design uploaded, proceeding with ABtest with different visitors
-
By 50%:50.5% we conclude both have similar favorit
Kurt, Data Analyst, Twitter:
Statistic in Data Analysis is really useful
Goes back to hundreds years ago when the they want to make some reference from the data
-
Because they make the data in the same page, and can make better analysis on the overall data.
-
Different data will have different results, and can't make data look good or the other way around
-
In the case earlier where people choose blue, we have selected one population who favorites blue
