Download presentation
Presentation is loading. Please wait.
1
Unsupervised Learning and Autoencoders
with Daniel L. Silver, Ph.D. Christian Frey, BBA April 11-12, 2017 10/11/2018 Deep Learning Workshop
2
Unsupervised Learning
A type of machine learning algorithm used to draw inferences from datasets consisting of input data without labeled responses Most common unsupervised learning method is cluster analysis, which is used for exploratory data analysis to find hidden patterns or groups in data K-Means is a classical method We will review a method called RBMs We can do unsupervised learning with supervised BP neural networks Called autoencoders … 10/11/2018 Deep Learning Workshop
3
Autoencoders using BP ANN
10/11/2018 Deep Learning Workshop
4
Autoencoders using BP ANN
Learn to predict input at the output Develops an efficient signal compressor Learn an internal representation (encoding) of image Creates feature detectors of the input at the hidden nodes Can judge similarity of input based on hidden vector similarity Can be used to pre-train a classifier 10/11/2018 Deep Learning Workshop
5
Autoencoders can be used to create semantic topologies
network Compress frequency of top 100 keywords of each document to just two hidden nodes Use different colours to represent the class of each document in 2D space Label by collour 10/11/2018 Deep Learning Workshop
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.