/Normalizations_Using_Menu_Driven

We have implemented Different types of normalizations techniques min-max normalization, Z-scale normalization, and decimal scaling normalization using menu- driven approach. We have used a user-input method for taking inputs from the user and perform normalization on the data value.

Primary LanguageJupyter Notebook

Normalizations_Using_Menu_Driven

We have implemented Different types of normalizations techniques min-max normalization, Z-scale normalization, and decimal scaling normalization using menu- driven approach. We have used a user-input method for taking inputs from the user and perform normalization on the data value.

What is Normalization

Data normalisation is a common machine learning technique that involves converting numeric columns to a standard scale. Some feature values differ from others multiple times in machine learning. The learning process will be dominated by features with higher values. As a result, normalising these values between scales is critical.

Types of Normalization

There are mainly three types of normalization techniques:

  1. min-max normalization: To convert data value between a specified new minimum and maximum value. The user need to specify old minumum, old maximum, new minimum and new maximum values to perform the normaalization. Let v denotes old data value and v' denotes corresponding new data value, then:

image

  1. Z-score normalization: The z-score method (also known as standardisation) converts data into a distribution with a mean of zero and a standard deviation of one.

image

  1. Decimal Scaling: It normalizes the data value between the range 0 to 1.

image

Python code implementation

We have used only simple menu driven if-else looping to perform the all normalization techniques.