Outliers

❮ Previous Next ❯

As we learn earlier that the mean helps us to quickly understand the different parts of the data.
However, the mean is highly sensitive to some specific values in our dataset.
These specific values are significantly different from others making them outliers as these don't fit within the majority of the dataset.
It is an essential step to identify these outliers, especially from the data science point of view. If they go unnoticed, they can cause errors in our data collection analysis.
Once we're able to spot outliers, we can decide whether or not they actually indicate a considerable, but a real deviation from the mean or whether they were caused by a sampling error.
Suppose we are handling a dataset of heights of some kids in inches, but somehow one of the kid's height's was recorded in centimeters, and the data set looks like this:
```
[50, 50, 51, 49, 48, 145]
```
In this case, the outlier is 145.
What if one of the kid's height is significantly higher than the other's:

[50, 50, 51, 49, 48, 64.5]

In this case, the outlier is 64.5.

❮ Previous Next ❯

Numpy Tutorials | Projects | Interview Questions Tutorial