Deep Learning - Unsupervised Pre-Training

Video Description

This video on focuses on the problem of training neural networks based on underfitting and overfitting, whereas unsupervised learning can be used for better regularization for overfitting. Also the video, is closely related to the field of machine learning as the unsupervised learning algorithm is a machine learning technique, “used to draw inferences from datasets consisting of input data without labeled responses” (MATLAB 1).

Other examples of unsupervised learning methods such as clustering analysis, “which is used for exploratory data analysis to find hidden patterns or grouping in data”, can be found here (MATLAB 2).





Sources Listed Below:

"Unsupervised Learning." Unsupervised Learning - MATLAB & Simulink. The MathWorks, Inc., n.d. Web. 25 Jan. 2017. <https://www.mathworks.com/discovery/unsupervised-learning.html>.

  • AI2 Conceptual Assignment: Machine Learning Videos  Neural networks [7.3] : Deep learning - A short video on the basic idea of unsupervised pre-training.