What is Classification and Regression?
Classification and Regression models are powerful tools to help users make predictions. The objective of classification models is to predict categorical values whereas regression models predict numerical values. The utilization of these tools can range from simply predicting patient wait times to diagnosing patients with cancer using digital MRI images.
Classification and Regression models are powerful tools to help users make predictions. The objective of classification models is to predict categorical values whereas regression models predict numerical values. The utilization of these tools can range from simply predicting patient wait times to diagnosing patients with cancer using digital MRI images.
Classification Demonstration
The example to the right is from the publicly available data set called “BreastCancer”. There are 699 observations with 11 predictor variables. The objective of the model is to train the computer how to identify which cells are “benign” and which cells are “malignant”. Questions to Ask:

Classification Plot
Decision boundaries is a graphical representation of the logic that partitions the plot into classification buckets. These buckets are represented by the colored lines and shadings. Each machine learning algorithm/formula develops a unique decision boundary. In the plot to the left, you will see the two classifications buckets of benign (red) and malignant (blue). The objective of a classification model is to build a decision boundary that aligns the observations within their respective decision boundaries (i.e. the blue dots within the blue decision boundaries and the red dots within the red decision boundaries). The accuracy of the model is equal to the percent of observations that are correctly classified into their respective buckets. 