Machine Learning

advertisement
Rise of the Machines
Jack Concanon
“A Field of study that gives computers the
ability to learn without being explicitly
programmed
- Arthur Samuel
“A computer program is said to learn from
experience E with respect to some class of tasks T
and performance measure P, if its performance at
tasks in T, as measured by P, improves with
experience E
- Tom M. Mitchell
Can Machines Think?
 Can machines do what we do?
 Can they pretend?
 How much help do they need?
Why Now?
http://internet-map.net
Supervised Learning
 Requires labelled training
examples
 Function approximation
 Based on an input vector
you should receive an
expected output
 Linear regression is the
simplest
 Line of best fit is a valid
tool for prediction
Unsupervised Learning
 Unclassified data
 Finding hidden structures
 Clustering
 Data mining
Linear Regression Example
Neural Networks
Neural Network Visualisation
Example – Cat or Hat?
Thanks for Listening!
•
•
•
•
•
•
How much data is made each minute? http://www.visualnews.com/2012/06/19/how-muchdata-created-every-minute/
What does a hidden layer do? https://stats.stackexchange.com/questions/63152/whatdoes-the-hidden-layer-in-a-neural-networkcompute#63163
Animating neural networks using R http://beckmw.wordpress.com/2013/03/19/animatingneural-networks-from-the-nnet-package/
Google research document on identifying high level
features from images of humans and cats –
https://static.googleusercontent.com/external_content/un
trusted_dlcp/research.google.com/en/us/archive/unsuper
vised_icml2012.pdf
Google research documents on machine learning http://research.google.com/pubs/ArtificialIntelligencean
dMachineLearning.html
Coursera Machine Learning https://www.coursera.org/course/ml
Download