Apply algorithm activity catalog

After the dataset has been transformed using preprocessing activities, the next phase is the application of a machine learning algorithm. The input of the transformed dataset and the algorithm into the Train Model activity produces the trained model.

Different algorithms accomplish different tasks. Algorithms examine the data and determine a model that is the closest fit to the data that is being reviewed.

The Algorithms activity catalog provides supervised and unsupervised algorithms:

  • Supervised algorithms are for regression, classification, or forecasting problems.
    • XGBoost
    • Linear Learner
    • Random Forest
    • Decision Tree
    • Extra Trees
    • Multilayer Perceptron
    • DeepAR: forecasting
  • Unsupervised algorithms are for cluster and associations analysis problems and dimensionality reduction.
    • K-Means
    • Auto Clustering
    • PCA: Principal Component Analysis
  • Custom algorithms are for packaging and deploying custom algorithm code to Coleman AI Platform. The code is used for model training.
Related topics