Scientist or engineer do many empirical experiments to gather data. Then they will use mathematical tools to do curve fitting analysis in order to derive the mathematical equation model that captures the characteristics or pattern of the behavior of subject under observation. Once they are satisfied that the discovered equation can predict the expected outcome given the known input then they will publish their empirical study and curve fitting equation for peer review.

But curve fitting mathematical tools is good for visualizing two dimensions variables (X, Y). If the variables are more than 3 it very hard for our human brain to draw or visualize the data points. Did our mind stop solving how to represent many variables problem? No, it turns out machine learning can be another mathematical tool to visualize many variables. We called this mathematical concept as latent space.

INSTRUCTION ON HOW TO USE GEOGEBRA:

Step 1: Delete text1 by selecting it and then selecting ⊠
Step 2: Delete conic equation labeled "c" and "d" and then selecting ⊠
Step 3: Enter in input as is: f(x)=5(e^(-x)) press enter button in your keyboard

Instruction on how to edit using Desmos Graphing: Click the lower right icon "edit graph on desmos" created by Sam Ortega 9/3/2017
Credit to: Desmos, Inc

Using Desmos graphing, graph the following analog signal:
f(t)=cos 100 π t + sin 200 π t + sin 60π t
Credit to: Desmos, Inc