https://en.wikipedia.org/wiki/Outline_of_statistics

Terms

sample, unbiased (representative) sample

sample mean, average $\bar x$

sample covariance, correlation

point estimate, confidence interval at confidence level $\gamma$ (also called confidence coefficient)
the higher level $\gamma$ we want, the wider the confidence interval we get

https://en.wikipedia.org/wiki/Type_I_and_type_II_errors: type I error is false positive (rejection of true null hypothesis), type II error is false negative (failure to reject a false null hypothesis)

Linear regression

and non-linear regression, quadratic function is the simplest example

Stohastic gradient descent

Least squares square method involves operations on matrices which can be huge.

Mini-batch SGD.

https://stats.stackexchange.com/questions/160179/do-we-need-gradient-descent-to-find-the-coefficients-of-a-linear-regression-mode
https://stats.stackexchange.com/questions/23128/solving-for-regression-parameters-in-closed-form-vs-gradient-descent

http://wiki.fast.ai/index.php/Gradient_Descent