Online-Academy
Look, Read, Understand, Apply

Data Mining And Data Warehousing

Predictive and Descriptive Data Mining

Predictive Mining

To predict unknown or future values based on the patterns found in the data.

How it works:
Uses historical data to build models that can forecast outcomes. Applies techniques like classification, regression, and time series analysis.
Examples:
Predicting customer churn in a telecom company.
Forecasting next month’s sales for a product.
Determining the likelihood that a loan applicant will default.

Techniques for performing predictive data mining are:
Decision trees
Neural networks
Support Vector Machines (SVM)
Linear/logistic regression

Descriptive Mining

Is done to find human-interpretable patterns that describe the data.

How it works:

Summarizes the main characteristics of the data, without trying to predict future outcomes.Focuses on understanding and explaining what's in the dataset.
Examples:
Finding customer segments with similar purchasing behavior.
Discovering association rules like "people who buy bread often buy butter."
Identifying frequent patterns in transaction data.
Techniques:
Clustering
Association rule mining
Summarization
Principal Component Analysis (PCA)