Machine Learning: Understand Centering and Scaling purposes
<p>This article introduces the centering and scaling concepts.<br />
With a real-world use case, I explain the advantages of the center and scale the data.</p>
<p>We dive into simple calculations and explanations by looking at Scikit-Learn ready-made methods.</p>
<p>Technically, we compare the MinMaxScaler, StandardScaler, and RobustScaler. They are part of transformers’ methods facilitating the preprocessing.</p>
<p>By the end, you will understand the purpose of centering and scaling data and be ready to use ready-made Scikit-Learn transformers.</p>
<h1>What is Centering and Scaling?</h1>
<h2>Understand the concepts</h2>
<p>Scaling transforms data to a specific range or scale, while centering involves shifting the data points so that their mean becomes zero.<br />
An example below.</p>
<p><img alt="" src="https://miro.medium.com/v2/resize:fit:700/1*gdID757053ZvQf6Wpwhdmg.png" style="height:342px; width:700px" />Image by </p>
<p><a href="https://medium.com/u/f03b470a2638?source=post_page-----f27155ef5fa--------------------------------" rel="noopener" target="_blank">Flo</a></p>
<p>You can see the effect of scaling and centering the data. Now, at the right, data are <strong>centered around 0</strong> and appear on a <strong>shorter scale</strong> (axes X and Y).</p>
<h2>Advantages</h2>
<p>There are several advantages to center and scaling the data, the most important for scaling</p>
<p><a href="https://pub.towardsai.net/machine-learning-understand-centring-and-scaling-purposes-f27155ef5fa">Read More</a></p>