Class notes
Week 1 Review
- Institution
- Arizona State University
Lecture notes of 6 pages for the course CSE 571 Artificial Intelligence at Arizona State University
[Show more]
Seller
Follow
Content preview
Week 1 ReviewMachine Learning
A major goal of a machine learning is achieving high generalization
Supervised learning (Learning through demonstration)
inputs and outputs are provided
build a model that learns the relationships between inputs and
outputs
Classification (discrete)
object recognition (class label (image is a banana, a toy, a mug,
etc.), diabetes classification
Age or weight may be used as individual inputs in a medical
classification problem. In machine learning, the term ‘feature’ is
often used to describe these individual inputs.
Regression (continuous)
stock predictions, weather forecasts, fuel consumption, airflow
around a plane wing, etc.
Unsupervised learning
Only inputs are provided, no labels
Best suited for learning spatial relations of a dataset
Clustering
uses distance between points and groups points based on
distances
Types of clustering
hard clustering, each point is assigned to a single group
soft clustering uses probability. How likely is it that a point
belongs to a particular class?
Week 1 Review 1
, In soft clustering, a sample can belong to multiple groups
to various degrees.
hierarchical clustering, class 1 and class 2 are both in class 3
Dimensionality Reduction
Manifold learning
The distances will likely differ between two samples in a reduced
dimension space on the manifold as compared to the original
space because the geodesic distance is not necessarily the same
as the Euclidean distance.
Reinforcement learning
Agent repeatedly interacts with the environment
learning though trial and error (fewer trials = more efficiency)
reward is provided by a reward function (challenging to specify)
a reward function takes the current state as input
Goal of reinforcement learning is to learn a policy that maximizes the
sum of rewards
A policy(π ) is a mapping that takes in the current state and yields
the appropriate action
Robotics Examples: Learning to lift an object, Throwing a ball into a
hoop
Challenges: Sample efficiency, especially in cases where there is a lot
of redundant learning. In cases where there are sparse rewards, it can
take a while before a model receives any type of reward.
Mean Square Error (loss function)
Square the distance from the model line to the data points so that negative
and positive differences do not cancel each other out.
The difference between the true output and the model output
1
N
∑N ^i )2
i=1 (y − y
i
Week 1 Review 2
The benefits of buying summaries with Stuvia:
Guaranteed quality through customer reviews
Stuvia customers have reviewed more than 700,000 summaries. This how you know that you are buying the best documents.
Quick and easy check-out
You can quickly pay through credit card or Stuvia-credit for the summaries. There is no membership needed.
Focus on what matters
Your fellow students write the study notes themselves, which is why the documents are always reliable and up-to-date. This ensures you quickly get to the core!
Frequently asked questions
What do I get when I buy this document?
You get a PDF, available immediately after your purchase. The purchased document is accessible anytime, anywhere and indefinitely through your profile.
Satisfaction guarantee: how does it work?
Our satisfaction guarantee ensures that you always find a study document that suits you well. You fill out a form, and our customer service team takes care of the rest.
Who am I buying these notes from?
Stuvia is a marketplace, so you are not buying this document from us, but from seller 4point0. Stuvia facilitates payment to the seller.
Will I be stuck with a subscription?
No, you only buy these notes for $2.99. You're not tied to anything after your purchase.
Can Stuvia be trusted?
4.6 stars on Google & Trustpilot (+1000 reviews)
50064 documents were sold in the last 30 days
Founded in 2010, the go-to place to buy study notes for 14 years now