Quantile Transformers
Quantile transformation techniques are the type of feature transformation technique that can be applied to NY numerical data observations. This transformation technique can be implemented using sklearn.
In this transformation technique, the input data can be fed to this transformer where this transformer makes the distribution of the output data normal to fed to the further machine learning algorithm.
Here there is a paramere called output_distribution, which value can be set to uniform or normal.
Python3
from sklearn.preprocessing import QuantileTransformer quantile_trans = QuantileTransformer(output_distribution='normal') data_transformed = quantile.fit_transform(data) |
Feature Transformation Techniques in Machine Learning
Most machine learning algorithms are statistics dependent, meaning that all of the algorithms are indirectly using a statistical approach to solve the complex problems in the data. In statistics, the normal distribution of the data is one that a statistician desires to be. A normal distribution of the data helps statisticians to solve the complex patterns of the data and gain valuable insights from the same. But for the algorithm scenario, a normal distribution of the data can not be desired every time with every type of dataset, which means the data which is not normally distributed needs preprocessing and cleaning before applying the machine learning algorithm to it.
In this article, we will be discussing the feature transformation techniques in machine learning which are used to transform the data from one form to another form, keeping the essence of the data. In simple words, the transformers are the type of functions that are applied to data that is not normally distributed, and once applied there is a high of getting normally distributed data.
There are 3 types of Feature transformation techniques:
- Function Transformers
- Power Transformers
- Quantile Transformers