Data Mining Processes

Solving data mining problems using an automated process is an ongoing dilemma. Hence, I have designed several data mining flows in order to accomplished this goal. The following are examples of data mining flows that I have designed:

Flow 1 illustrates the process of loading the training set into SimpleKMeans clusterer. Then, the predictive model gets built using J48 based on the generated clusters. All this process is evaluated using cross-validation. In this flow, FilteredClassifier contains both cluster and classification schemes.

Flow 2 allows evaluating the learning algorithm using 10-fold cross-validation method.

Flow 3 represents classification process using J48 algorithm. It provides attribute summary, displays training/test sets (training set data is 66%), and visualizes the resulting tree model. Finally, the flow displays classification result.