Unsupervised Learning

 In unsupervised learning, a model is trained on unlabelled data, and try to find patterns in the data. Unsupervised learning helps to uncover patterns in the data that you are unaware.

 

Example

1. Cluster customers into groups based on their past data with us.
2. Group documents by their content
3. Group news articles by their content
4. Recommend movies/tv shows to the users
5. Group products based on their similarities
6. Detect fraudulent transactions

 

Since Unsupervised learning doesn't require any labelled data, it can  detect new and unknown fraud types from the unknown data, and prevent the damages before it happens.

 

Unsupervised learning vastly used in following tasks.

 

1. Clustering: Group the data points together based on some similarities.
2. Dimensionality reduction: Reduce the number of features in a dataset while preserving the important information.
3. Anomaly detection: Used to identify outliers. Outlier is a data point that is significantly different from the rest of the data points in a dataset.

 

key terms in unsupervised learning

 

1. Clustering: Grouping the data points based on the similarities.
2. Centroid: It represents center of the cluster.
3. Dimensionality Reduction: Reduce the number of input features while preserving important information.
4. Outlier: A data point that is significantly different from other data points, most of the times a outlier indicates an anomaly.
5. Entropy: Measure the impurity, uncertainty or randomness in a dataset. It is used to evaluate the quality of clustering algorithm. A clustering algorithm with high entropy will create clusters with high disorder, low entropy will create clusters with low disorder.
6. Curse of dimensionality: This situation arise, when number of features are more than the number of data points.

Previous                                                    Next                                                    Home