Which data mining technique is used to predict a numerical value given characteristics of each member of a dataset?

Which data mining technique is used to predict a numerical value given characteristics of each member of a dataset?

• A. Classification
• B. Regression
• C. Clustering
• D. Association rule learning

Answer: (B)

Predicting a numerical value from input features is what regression is designed for. Regression models learn how a continuous target variable changes with the characteristics of each item, producing a numeric prediction for new data. The training process uses actual numeric outcomes to fit a relationship between the features and the target, which can be linear or more flexible with nonlinear forms. This is different from classification, which assigns items to discrete categories; clustering, which groups similar items without a target numeric outcome; and association rule learning, which discovers patterns of co-occurring attributes rather than predicting a numeric value. In practice you’d assess regression performance with metrics like RMSE or MAE to gauge how close the predictions are to the true numbers.

Predicting a numerical value from input features is what regression is designed for. Regression models learn how a continuous target variable changes with the characteristics of each item, producing a numeric prediction for new data. The training process uses actual numeric outcomes to fit a relationship between the features and the target, which can be linear or more flexible with nonlinear forms. This is different from classification, which assigns items to discrete categories; clustering, which groups similar items without a target numeric outcome; and association rule learning, which discovers patterns of co-occurring attributes rather than predicting a numeric value. In practice you’d assess regression performance with metrics like RMSE or MAE to gauge how close the predictions are to the true numbers.