THE ARTISTS OF DATA SCIENCE PRESENT...
THE MATH YOU
ABSOLUTELY
NEED FOR
DATA SCIENCE
CURATED BY:
Harpreet Sahota
The Artists of Data Science
OVERVIEW
Yes, you need to know math to be in data science. But do you need to know all
of it to get your first job in data science? Absolutely not. What you do need is a
firm grasp on some fundamental concepts. If you understand these
fundamentals then learning anything else you need won't be too much of a
stress. Focus on the fundamentals, and know them intuitively.
WHAT I BELIEVE
I believe that any body can learn
the essential tools of mathematics
for data science. Just because you
didn't major in math or stats or
computer science doesn't mean
that you can't learn those skills. At
one point in my life I didn't know
how to walk. Now I walk all the
time. There was also a point when I
didn't know how to code, and now I
do it daily. You can learn anything.
OBJECTIVES
THE GOAL OF THIS PAPER
I'll start with the basics: Listing out everything I feel you need to know
before even thinking about applying for your first job in data science. I'll
link to my favorite resources so you can go more in depth. These are ones
that I have hand picked for their clarity and delivery of content.
KEY POINTS
This list is not exhaustive. There's gonna be some pedantic academic out
there talking out of mouth about:"Well you need to know this, and that,
and what about that". If you're that person, relax. Close this document
now. And go on about your day.
The skills outlined here are the foundations. They are the fundamentals
on which everything else is built. If you study these, know them well,
understand them intuitively, then you'll have no problem picking up
more complex concepts.
This is not a textbook. I will generously link to resources that explain
things far better than I can. They're not random resources. They are ones
that I fully co-sign.
Pretty much everything here is lifted from various university
courses. Mostly from Penn State. They're awesome.
ALGEBRA
Knowledge of the following mathematical operations is required for
data science:
Addition
Subtraction
Division
Multiplication
Radicals (i.e., square roots)
Exponents
Summations (∑)
Factorials (!)
Scientific Notation
When performing a series of mathematical operations, begin with
those inside parentheses or brackets. Next, calculate any exponents
or square roots. This is followed by multiplication and division, and
finally, addition and subtraction.
Parentheses
Exponents & Square Roots
Multiplication and Division
Addition and Subtraction
Take a self assessment here and check your results here.
CALCULUS
For long-term success in the field of data science, it is imperative
that you have a working knowledge of multidimensional calculus.
We'll do a simple review of the calculus techniques most frequently
used in white papers and research papers for statistics and machine
learning:
Summation and Series
Limits
Differentiation
Integration
Multivariate calculus
Learn more about The Binomial Theorem
CALCULUS - SERI ES
Learn more about convergence and
divergence of series' here
CALCULUS - SERI ES
Learn more about geometric series here
Learn more about the Taylor series here
CALCULUS - DERI VATI VES
The definition of a derivative is
The derivative is the slope of the tangent line to
the graph of f(x), assuming the tangent line exists.
Here are some common rules you should know:
Power Rule
Product Rule
Quotient Rule
Chain Rule
L'Hopital's Rule
Learn more about what the limit of a function is
here and here.
Learn more about derivaties here.
CALCULUS - I NTEGRALS
For a function f(x), its definite integral is:
Some other concepts related to integrals
you should be aware of:
The Fundamental Theorem of
Calculus
Integration using substitution
Integration by parts
Learn more on integrals here, and here.
CALCULUS - MULTI VARI ATE
Learn more about partial derivatives here and
here.
CALCULUS - MULTI VARI ATE
Integrating over regions is important in
calculus and has many applications in
probability theory. Suppose we have a
function f(x,y) over some region R
Learn more about double integrals here and here.
Visit here to learn how integrals are used in
probability theory.
LI NEAR ALGEBRA - MATRI CES
LI NEAR ALGEBRA - MATRI CES
LI NEAR ALGEBRA - MATRI CES
Learn more about the transpose of a
matrix here.
LI NEAR ALGEBRA - MATRI CES
Learn more about the matrix addition
here.
LI NEAR ALGEBRA - MATRI CES
Learn more about the matrix scalar
multiplication here.
LI NEAR ALGEBRA - MATRI CES
Learn more about the matrix
multiplication here.
LI NEAR ALGEBRA - MATRI X
PROPERTI ES
Learn more about the identity matrix
here.
LI NEAR ALGEBRA - MATRI X
PROPERTI ES
LI NEAR ALGEBRA - MATRI X
PROPERTI ES
Learn more about the identity matrix
determinants here.
LI NEAR ALGEBRA - MATRI X
I NVERSES
Learn more about matrix inversion here,
here, and here.
LI NEAR ALGEBRA - ADVANCED
PROPERTI ES
Learn more here.
LI NEAR ALGEBRA - ADVANCED
PROPERTI ES
Learn more about Euclidean norm here.
Learn more about norms here.
LI NEAR ALGEBRA - ADVANCED
PROPERTI ES
LI NEAR ALGEBRA - RANGE OF A
MATRI X
LI NEAR ALGEBRA - NULLSPACE
OF A MATRI X
Learn more about null space here.
LI NEAR ALGEBRA - OTHER
TOPI CS
Some other important topic
include:
Gauss-Jordan Elimination
Using Gauss-Jordan to Solve a
System of Three Linear Equations
- Example 1
Algebra - Matrices - Gauss Jordan
Method Part 1 Augmented Matrix
Gaussian Elimination
LI NEAR ALGEBRA -
EI GENDECOMPOSI TI ON
Learn about everything related to
Eigens here.
A concept especially important in PCA
is the Singular Value Decomposition.
Learn more about that here.
ADDI TI ONAL RESOURCES
All of the above can be reviewed below.
This is a list of some of my personal
favorite places that have helped me on
my journey:
PatrickJMT
NancyPi
Krista King
CatBug88
MathByFives
CLOSI NG REMARKS
There is a lot of math to learn. That is for
sure.
But don't feel like you have to memorize
everything. Go through the topics as
quickly as you can get gain an awareness
of the concepts and use this document as
a refresher for when you forget some stuff.
You won't have to solve problems by hand
on the job. But you will need to read
papers, mostly academic ones.
So it's a good idea to be aware of these
concepts so you can progress in your
career.
Thanks!
