Distinction between the terms Machine Learning, Pattern Recognition, and Classification.
Note: the distinction between machine learning and pattern recognition is not so clear and they are sometimes used as synonyms. There are disciplines that prefer the one or the other term for de fact the same thing. However, if there’s a distinction then machine learning is more general than pattern recognition.
Machine learning
Systems that can “learn” form experience.
- Supervised machine learning: system trained with known input/output pairs. System generalises this data so that it can predict output to unseen input.
- Unsupervised machine learning: system operates directly on inputs whose output is not known (no training) and attempts to discover structure in the data. Clustering is sometimes a synonym to unsupervised machine learning.
Pattern Recognition
Assignment of a label to a given input value. It’s a sub-branch of machine learning. The same distinction of supervised and unsupervised approaches as in machine learning holds here.
Classification
Assigning an unknown input to a class. It is one type of pattern recognition. Other types of pattern recognition are regression and parsing.
Machine Learning and Pattern Recognition Algorithms
Supervised
- Hidden Markov models
- Decision trees
- K-nearest-neighbours
- Naive Bayes classifier
- Artificial neural networks
- Support vector machines
- Genetic programming
Unsupervised
- Clustering
- Hidden Markov models
References
- http://en.wikipedia.org/wiki/Machine_learning
- http://en.wikipedia.org/wiki/Pattern_recognition
- http://www.quora.com/What-is-the-difference-between-pattern-recognition-and-machine-learning
- http://stackoverflow.com/questions/1832076/what-is-the-difference-between-supervised-learning-and-unsupervised-learning