Non-linear regression methods - where and how to use them - form a vast field in science.
A quick practical intro can be found here:
https://statisticsbyjim.com/regression/difference-between-linear-nonlinear-regression-models/
How to use non-linear regression methods in R:
http://www.sthda.com/english/articles/40-regression-analysis/162-nonlinear-regression-essentials-in-r-polynomial-and-spline-regression-models/
Just a note - an incorrect assumption that people often make when starting using regression methods to fit data is that a linear regression can be used to estimate only "linear curves" e.g.
https://upload.wikimedia.org/wikipedia/commons/thumb/3/3a/Linear_regression.svg/2880px-Linear_regression.svg.png
However, linear regression can be used to fit a large variety of curves to data as shown in the above articles, and also here:
https://upload.wikimedia.org/wikipedia/commons/thumb/8/8b/Polyreg_scheffe.svg/2880px-Polyreg_scheffe.svg.png
There are a set of assumptions (that are fairly general) describing the relationship between dependent and independent variables - if satisfied a linear regression is a good way to fit/model data:
https://en.wikipedia.org/wiki/Linear_regression#Assumptions
There are methods to test if these assumptions are satisfied in a given dataset - here is an intro on that:
http://people.duke.edu/~rnau/testing.htm
As an aside, also note that methods like PCA (e.g. we showed used for dimensionality reduction) also make assumptions that data can be "linearly transformed". An excellent primer on the PCA method is here:
https://www.cs.cmu.edu/~elaw/papers/pca.pdf
There are non-linear generalizations of PCA, e.g. kernel PCA:
https://en.wikipedia.org/wiki/Principal_component_analysis#Nonlinear_PCA
Please let us know if you have questions on this topic - it is fairly important, it quickly gets deeper in math, but there are methods that are accessible and easy to use w/o getting into the math details.