Unsupervised Learning


Unsupervised learning is a category of machine learning that does not use labeled data to predict outcomes, but instead relies on the computer to sort out patterns and similarities in the larger dataset.

  • is a subset of machine learning in which data scientists feed the model unlabeled data sets, asking it to identify patterns in data on its own, without guidance from the data scientist
  • algorithm does not know the correct answers
  • looks for patterns, rather than using predefined inputs/outputs as with supervised learning
  • can uncover hidden groupings or patterns in the data
  • uses only the raw data for the input , and we supply no

Clustering

Clustering is an unsupervised machine learning technique used to group a set of data points into clusters or groups in order to identify natural existing patterns within a dataset.

  • organizes objects into groups whose members are alike in some way
    • without a training set
  • goal is to categorize data points into discrete clusters
    • the objects within are distinct from, yet similar to, those in other clusters
  • different ways to cluster data:
    • k-means clustering
  • can use scikit-learn Python library for machine learning