You are on page 1of 27

MACHINE LEARNING

101
AGENDA
o What is Machine Learning?
o Machine Learning Algorithms
o Machine Learning Process
o Demo
ML ?
ML ?
Machine learning at its most basic
is the practice of using algorithms
to parse data, learn from it, and
then make an inference about
something in the world - Nvidia
MACHINE LEARNING
Learn from Experience

Learn from Experience Follow Instructions


TYPES OF

ML
TYPES OF

ML
• SUPERVISED LEARNING

• UNSUPERVISED LEARNING

• REINFORCEMENT LEARNING
SUPERVISED LEARNING
• Takes data that already has an answer (labeled
data)

• For an example:
o Classify images that has oranges in it
o We need images that have already been
labeled as an orange or not an orange

• Its supervised because we can tell the model


what it got wrong
Training Dataset Testing Dataset

Image Label Image Label

True ?
True ?
False ?
True ?
False ?
SUPERVISED LEARNING
• Takes data that already has an answer (labeled
data)

• For an example:
o Classify images that has oranges in it
o We need images that have already been
labeled as an orange or not an orange

• Its supervised because we can tell the model


what it got wrong
ALGORITHMS
Supervised learning
CLASSIFICATION ALGORITHMS
• Classification is about predicting labels

• Examples:
 Is this an orange?
 Is this an apple, orange or lime?

• Classification algorithms:
 Linear Classifiers
 Support Vector Machines
 Decision Trees
 Boosted Trees
 Random Forest
REGRESSION ALGORITHMS
• Regression is about predicting quantity

• Examples:
 How much is the house?
 What is the salary?
 What is the end of the day
balance?

• Regression algorithms:
 Linear Regression
 Logistic Regression
 Decision Trees
 Naïve Bayes Classification
ML PROCESS
DEMO
QUESTIONS?
EXTRA
UNSUPERVISED LEARNING
• No supervision

• Takes data that has not been labeled

• Based on the hidden patterns and structures of


the data, it creates clusters

• Its unsupervised because there's no way of


telling the model what's wrong or right and we
let the model do its own thing
UNSUPERVISED LEARNING
• No supervision

• Takes data that has not been labeled

• Based on the hidden patterns and structures of


the data, it creates clusters

• Its unsupervised because there's no way of


telling the model what's wrong or right and we
let the model do its own thing
REINFORCEMENT LEARNING
• It involves an autonomous agent learning how to
navigate an uncertain environment with the goal
of maximizing a numerical reward

• Examples:
• Self driving cars
• Google Deep Mind’s AlphaZero was able
master the games of chess, shogi (Japanese
chess), and Go, and beat world-champion in
each case
Reference
• Machine Learning For Beginners From Zero Level
https://blog.usejournal.com/machine-learning-
for-beginners-from-zero-level-8be5b89bf77c

• A Friendly Introduction to Machine Learning


https://www.youtube.com/watch?v=IpGxLWOI
Zy4&list=PLEr9REFgKb-
N0JaIvO1flOCD450XO6sat&index=5&t=0s
• What’s the Difference Between Artificial
Intelligence, Machine Learning, and Deep
Learning?
https://blogs.nvidia.com/blog/2016/07/29/wha
ts-difference-artificial-intelligence-machine-
learning-deep-learning-ai/
• Fusion Alliance Machine Learning 101 Deck

You might also like