Machine learning analysis

Machine learning in Analytics is automated (AutoML). Complex computational work such as data preprocessing, algorithm selection, hyperparameter tuning, and model validation is performed for you by Analytics. This automation allows you to put machine learning to work on company data with relatively little effort, and without requiring that you have specialized data science capabilities.

Supervised and unsupervised machine learning

Analytics supports both supervised and unsupervised machine learning.

Supervised machine learning uses existing data labeled with categories or numeric values as the basis for predicting categories or numeric values in similar, unlabeled data.

Unsupervised machine learning discovers categories in uncategorized or unlabeled data.

Machine learning not supported on 32-bit computers

If you install Analytics on a 32-bit computer, the machine learning operations are not supported, and the Machine Learning menu does not appear. The computation required by machine learning is processor-intensive and better suited to 64-bit computers.

Machine learning operations

Operation

ML type

Supported data types

Functionality

Output
Train Supervised

Character

Numeric

Datetime

Logical
  • Uses automated machine learning to create a predictive model.
  • Predictive model file

    (*.model)

  • Model evaluation file

    (Analytics table)
Predict Supervised

Character

Numeric

Datetime

Logical
  • Applies a predictive model to an unlabeled data set to predict classes or numeric values.
  • Results file

    (Analytics table)
Cluster Unsupervised Numeric
  • Groups numeric data.
  • Groups records based on similar, or nearby, values in one or more numeric fields.
  • Results file

    (Analytics table)