Teradata Warehouse Miner

Teradata Warehouse Miner (TWM) is a software application provided by Teradata for achieving data mining inside RDBMS databases without the need for migrating the data from the data warehouse.^[1] TWM is bundled with multiple statistical and machine learning algorithms for analyzing data and creating data mining models. The whole data mining process, e.g. CRISP-DM process, is supported using TWM which includes: descriptive statists and analysis, analytic data set creation, data mining model creation and data mining model scoring.^[2] This facilitates Knowledge Discovery in Databases (KDD) which provides business intelligence for organisations.

Overview

Teradata Warehouse Miner is a software application bundle created by Teradata for in-database data mining. It supports running on RDBMS databases using ODBC connections.^[3] It supports multiple algorithms for data mining like K-means clustering, association rule mining, decision tree analysis and regression. It also supports multiple statistical analysis commonly found in statistical applications.^[4]

Supported data analysis & data mining algorithms

• Descriptive Statistics
  • Frequency
  • Histogram
  • Scatter plot
  • Data Explorer
• Data Transformation & Data Reduction
  • Variable creation
  • Variable transformation
  • Data reduction matricies
  • Analytic Data Set creation
• Statistical Tests
  • Binomial test
  • Chi-squared test
  • Rank tests (e.g. Mann-Whitney)
  • Kolmogorov–Smirnov test
  • Parametric (ANOVA tests)
• Data Mining
  • Factor Analysis
  • Regression (Linear and Logistic regressions supported)
  • Decision Tree
  • Clustering
  • Association Rules
  • Neural Networks (using a TWM Add-In)

See also

• Teradata
• Data Warehousing
• Data Mining
• Statistical analysis

References