-
EDA: Exploratory Data Analysis
-
ubiquitous: can explain everything from data
-
Data and Analysis is directly correlated. We have to someway analyze our data
-
By doing analysis we can make predictive model
-
Have insight and take good question about our data.
-
Defend agains bad data, wrangling, and cleaning the data.
-
Make a better visualization to present to larger part of audience.
-
Data Analyzis can benefit as follows:
-
Win competition
-
Has data reasoning
-
Improve communication
-
and extend our carreer.
-
investigate what's interesting
-
dig what's more interesting
-
The goal is have an impact so the world have an understand better.
-
how to represent mindset.
-
Aude once use conditional probability from given hometown, what is the probabilty of they're moving to another city.
-
Then she found out that most likely, we still live in our hometown
-
But in certain cases, things are different. So she investigate it with different approach.
-
-
contiditionalprob iven hometown what people in particular city.
-
city most likely still your homewtown.
-
certain cases, have different high prob. she curios and take different approach
-
She is fascinating to see pattern particular cities, in some countries, not necesssary expecting that.
-
She saw that the distribution of people living in same city is very flat.
-
When looking at the data, we may want to choose at the world scale, to give better intuition.
-
She back then only choose US data, but take different approach by observe world scale.
-
develop intuition about how data exists
-
By doing that we can select which features is the most significant, and what statistical tool to build our predictive model.
-
In conducting EDA, always be skeptical as we will develop new instuition.
-
Be open mind in detecting data, and have summary of the data.