## DATA MINING APPLICATION USING DECISION TREE: SURFACE FINISH ANALYSIS(part2)

Data Acquisition

The parameters considered are Spindle Speed, Feed Rate, and Depth of Cut According to the acceptable ranges of cutting speed, feed rate, and depth of cut when cutting 6061 aluminum one inch cubic block with a high speed steel cutter four levels of spindle speed – 750, 1000, 1250, and 1500 revolutions per minute (rpm), seven levels of feed rate – 6, 9, 12,15, 18, 21 and 24 inch per minute (ipm), and three levels of depth of cut – 0.01, 0.03, and 0.05 inch (in) were determined. The surface roughness (Ra) was measured in micro inches (min) by a stylus-based profilometer.

Decision Trees

Decision trees are simple and successful predictive learning algorithms. Learning using decision trees consist of two steps. In the first step, a tree is constructed using the training data. Then, for each record, the tree is traversed to determine the class to which the record belongs. Each internal node in the tree indicates a test on an attribute, each branch represents an outcome of the test, and each leaf node pointâ€™s class DT algorithms can be used to test conditional independence relations among variables in large data sets.

Decision tree model

CASE 1: IF Ra VALUE IS 0 to .00263 THEN Ra IS NORMAL

CASE 2: 750<=RPM<=1500, .254<=DOC<=.762, 152.4<=FR<=304.8

CASE 3: 750<=RPM<=1500, .762<DOC<=1.27, 304.8<FR<=609.6

Artificial Neural Networks

Artificial neural networks (ANN) are modelling techniques derived from human brain. They can be used to model complex non-linear relationships between inputs and outputs or to find patterns in data. An ANN consists of interconnected processing elements, nodes, where every connection has a weight. As decision trees, ANN approach requires a graphical structure to be built before applying it to the data. General topology of an ANN is shown in Figure