Machine learning is used…
- for database mining (web analytics coupe with user behavior, medical records)
- to develop programs humans can’t do by hand (aka: get the computer to learn and create its own program to do a task)
- to run self-customizing programs (aka: the reason Netflix “knows” which show or movie you’ll probably enjoy)
- to understand human learning (the brain and real AI)
Learning algorithms come in a few forms:
- supervised learning
- unsupervised learning
- reinforcement learning
- recommender systems
Supervised and Unsupervised learning algorithms are the most common. More about those two…
- the algorithm is given a dataset in which the “right answers” are given for specific constraints.
- can be also a regression problem as the algorithm is run to predict a continuous valued output
- can be a classification problem, where you attempt to predict a discrete valued output. It can have multiple outputs.
- dataset is provided without a structure – the algorithm is used to find the structure or relationships and sort the data accordingly.
- common problem, the Cocktail Party Algorithm: [W,s,v] = svd((repmat(sum(x.x,1),size(x,1),1).x)*x’);
Unsupervised learning algorithms are used in Social Network Analysis! Lovin’ the overlap with my other course.
Above are notes I’ve taken to keep track of the material covered in the Machine Learning course I am taking on Coursera. Notes are derived from week-based video lectures and supplemental information I’ve found online to help clarify certain concepts. If you are taking this course, most of the information below will be found in Week 1 associated material. If you are not taking the course, I hope you find the information below helpful in your search