Tuesday, 8 November 2016

Difference between supervised, unsupervised Machine learning

Difference between supervised, unsupervised Machine learning:

As we studied about the introduction of Machine learning in the previous post, I am just refreshing again in two lines below.

Machine learning is a field of computer science, probability theory, and optimization theory; this allows the people complex tasks to be solved for which a logical/procedural approach would not be possible or feasible.

The different categories of machine learning are shown below
  • Supervised learning
  • Unsupervised learning

Supervised learning:

In supervised learning, we have some really complex function from inputs to outputs. we have lots of examples of input/output pairs, but we don't know what that complicated function is. Using supervised learning algorithm we can make it possible.

  • Let us assume given a large data set of input/output pairs, to predict the output value for some new input value that you may not have seen before. 
  • We have to write the basic method that we need to divide the data set down into training set and a test set.
  • You have some model with an associated error function which you try to minimize over the training set, and then you make sure that your solution works on the test set. Once you have repeated this with different machine learning algorithms and/or parameters until the model performs reasonably well on the test set, then you can attempt to use the result on new inputs.
  • In this case the changes are done in model, which is data, but we do not need to change the program.

So, the supervised learnig deals with the classification.

The below are the algorithm types:
Linear Classifiers
  1. Logical Regression
  1. Naïve Bayes Classifier
  1. Perceptron
  1. Support Vector Machine
Quadratic Classifiers
K-Means Clustering
Decision Tree
  1. Random Forest
Neural networks
Bayesian Networks

Unsupervised Learning:   

  • Unsupervised learning is the other type of machine learning, which is used to draw inferences from datasets consisting of input data without labeled responses. 
  • In the unsupervised learning the most useful method is cluster analysis.

o Cluster analysis: This analysis is used for exploratory data analysis to find hidden patterns or grouping in data. The clusters are modeled using a measure of similarity which is defined upon metrics such as Euclidean or probabilistic distance.

The common clustering algorithms include.

  1. k-Means clustering: This algorithm partitions data into k distinct clusters based on distance to the centroid of a cluster
  2. Gaussian mixture models: These models clusters as a mixture of multivariate normal density components
  3. Hierarchical clustering: This algorithm builds a multilevel hierarchy of clusters by creating a cluster tree
  4. Hidden Markov models: This algorithm uses observed data to recover the sequence of states
  5. Self-organizing maps: This algorithm uses neural networks that learn the topology and distribution of the data.

Unsupervised learning methods are commonly used in the below areas.

  1. Bioinformatics: used for sequence analysis and genetic clustering;
  2. Data Mining: used for sequence and pattern mining; in medical imaging for image segmentation; 
  3. Computer vision: mainly used for object recognition.

No comments: