I Algorithms!

Welcome to you print-at-home Spread the love...

version of the I Love Algorithms Please use this deck, share it,
card deck. Simply print at home and credit us. We would like to
(color is best!) and cut out the know how you use it. For more
cards to begin playing with info visit dschool.stanford.edu.
algorithms. This deck can be
used as-is, on its own, or go to
https://dschool.stanford.edu/
emerging-tech to pair with
other tools we're building.

I Algorithms! I Algorithms!
I love...what? So...why do I need to know
An algorithm is a piece of code anything about algorithms?
that helps process, manipulate, or In order for the technologies of
investigate a data set. There are today and tomorrow (and all of
many types of algorithms used in the things that they power) to
machine learning. Machine learning represent all of us, they need to be
is a way of analyzing big data sets built by all of us. You don’t need
where the computer learns and gets to be the coder, but you need to
better at its analysis with time. know what the code can do.
This card deck describes six different If you understand what machine
machine learning algorithms. learning algorithms can do, you can
Since humans all learn differently, better envision the implications of
we show each algorithm in three your designs. You can influence
ways: (1) text description, (2) conversations about data and bias.
cartoon, and (3) types of questions
you might be asking yourself.

I Algorithms! I Algorithms!
Can my love run deeper
than these six algorithms? We hope you love this deck as
Of course. This deck includes much as you love algorithms!
the basic types of machine This is the work of the Stanford
learning algorithms, but there d.school. The d.school helps
are many ways to expand on this people develop their creative
content in future versions. abilities. It’s a place, a community,
One concept we have not covered and a mindset. As designers and
here is deep learning. Deep learning is educators, we believe in providing
a type of machine learning where the radical access to the intersection
computer learns what to do without of technology and design.
being explicitly programmed to do so. Spread the love...
It is the gateway to things like speech Please use this deck, share it,
and image recognition, analysis and credit us. We would like to
of text, and much more. Neat! know how you use it. For more
info visit dschool.stanford.edu.
Reinforcement Reinforcement Reinforcement
Learning Learning Learning
How do I win this game?
Put your machine into an How might this car drive itself?
environment and give it a How to optimize marketing so
goal. It begins to interact and someone will click click click?
uses trial and error to figure
out what to do. It wants to
win more than anything.
Useful in robotics. Useful for
figuring out ideal behavior
in a given situation in order
to maximize performance.

Regression Regression Regression

For finding cause and effect
between different variables.
Useful for forecasting (like the
weather) or for things where
historical data helps predict
the future. Regression is
your trend-finder. Feed it the
data and example answers. How much will my tiny
It compares its answers with house on the flood plain
the right ones to get better. be worth in 2020?
Did someone really buy 18
inflatable swans or is that
a fraudulent transaction?

Dimensionality Dimensionality Dimensionality

Reduction Reduction Reduction

Reduces the number of

variables in a data set but
keeps the important stuff.
Good for raw data sets where
a lot of features might be
redundant or irrelevant. Helps
see the forest through the
trees. You give it the data
but the machine figures Can you just
out how to clean it up. tell me what’s
important in
my data!!?
Clustering Clustering Clustering

Groups similar things

together. Makes groups
where objects in one group
are more similar to each I need to
other than to a different market these
group. You give this the ripped jeans,
data, but the machine are there certain
figures out how it’s related. types of groups that I
should target specifically?
How are the the world’s
consumers of kombucha
related? What are the
sub-segments?

Association Association Association

Are certain things likely

to happen together? The
algorithm finds hidden If someone buys a donut,
relationships. You give it are they 99% likely to
the data, but the machine also get a coffee?
figures out how it’s related. If someone listens
to Queen and
Kanye are they
65% likely to
also listen to
Funkadelic?

Classification Classification Classification

Is this a picture of
This algorithm predicts what hot dogs or legs?
category something might Does this x-ray indicate the
land in. You (the human) patient has pneumonia?
supervise it. You give it the Is that a noun or a verb or
data and you tell it what an adjective?
categories to pick from. It can
compare its answers with the Am I in the
right ones and get better. car lane or
the bike lane?
I Algorithms!
Welcome to you print-at-home Spread the love...
version of the I Love Algorithms Please use this deck, share it,
card deck. Simply print at home and credit us. We would like to
(color is best!) and cut out the know how you use it. For more
cards to begin playing with info visit dschool.stanford.edu.
algorithms. This deck can be
used as-is, on its own, or go to
https://dschool.stanford.edu/
emerging-tech to pair with
other tools we're building.

I Algorithms! I Algorithms!
I love...what? So...why do I need to know
An algorithm is a piece of code anything about algorithms?
that helps process, manipulate, or In order for the technologies of
investigate a data set. There are today and tomorrow (and all of
many types of algorithms used in the things that they power) to
machine learning. Machine learning represent all of us, they need to be
is a way of analyzing big data sets built by all of us. You don’t need
where the computer learns and gets to be the coder, but you need to
better at its analysis with time. know what the code can do.
This card deck describes six different If you understand what machine
machine learning algorithms. learning algorithms can do, you can
Since humans all learn differently, better envision the implications of
we show each algorithm in three your designs. You can influence
ways: (1) text description, (2) conversations about data and bias.
cartoon, and (3) types of questions
you might be asking yourself.

I Algorithms! I Algorithms!
Can my love run deeper
than these six algorithms? We hope you love this deck as
Of course. This deck includes much as you love algorithms!
the basic types of machine This is the work of the Stanford
learning algorithms, but there d.school. The d.school helps
are many ways to expand on this people develop their creative
content in future versions. abilities. It’s a place, a community,
One concept we have not covered and a mindset. As designers and
here is deep learning. Deep learning is educators, we believe in providing
a type of machine learning where the radical access to the intersection
computer learns what to do without of technology and design.
being explicitly programmed to do so. Spread the love...
It is the gateway to things like speech Please use this deck, share it,
and image recognition, analysis and credit us. We would like to
of text, and much more. Neat! know how you use it. For more
info visit dschool.stanford.edu.
Reinforcement Reinforcement Reinforcement
Learning Learning Learning
How do I win this game?
Put your machine into an How might this car drive itself?
environment and give it a How to optimize marketing so
goal. It begins to interact and someone will click click click?
uses trial and error to figure
out what to do. It wants to
win more than anything.
Useful in robotics. Useful for
figuring out ideal behavior
in a given situation in order
to maximize performance.

Regression Regression Regression

For finding cause and effect
between different variables.
Useful for forecasting (like the
weather) or for things where
historical data helps predict
the future. Regression is
your trend-finder. Feed it the
data and example answers. How much will my tiny
It compares its answers with house on the flood plain
the right ones to get better. be worth in 2020?
Did someone really buy 18
inflatable swans or is that
a fraudulent transaction?

Dimensionality Dimensionality Dimensionality

Reduction Reduction Reduction

Reduces the number of

variables in a data set but
keeps the important stuff.
Good for raw data sets where
a lot of features might be
redundant or irrelevant. Helps
see the forest through the
trees. You give it the data
but the machine figures Can you just
out how to clean it up. tell me what’s
important in
my data!!?
Clustering Clustering Clustering

Groups similar things

together. Makes groups
where objects in one group
are more similar to each I need to
other than to a different market these
group. You give this the ripped jeans,
data, but the machine are there certain
figures out how it’s related. types of groups that I
should target specifically?
How are the the world’s
consumers of kombucha
related? What are the
sub-segments?

Association Association Association

Are certain things likely

to happen together? The
algorithm finds hidden If someone buys a donut,
relationships. You give it are they 99% likely to
the data, but the machine also get a coffee?
figures out how it’s related. If someone listens
to Queen and
Kanye are they
65% likely to
also listen to
Funkadelic?

Classification Classification Classification

Is this a picture of
This algorithm predicts what hot dogs or legs?
category something might Does this x-ray indicate the
land in. You (the human) patient has pneumonia?
supervise it. You give it the Is that a noun or a verb or
data and you tell it what an adjective?
categories to pick from. It can
compare its answers with the Am I in the
right ones and get better. car lane or
the bike lane?