1The University of Iowa Intelligent Systems LaboratoryData Mining: STATISTICAThe University of Iowa Intelligent Systems LaboratoryOutline•Prepare the data•Classification and regression2The University of Iowa Intelligent Systems LaboratoryPrepare the Data• Statistica can read from Excel, .txt and many other types of files• Compared with WEKA, Statistica is much easier in terms of data preparingThe University of Iowa Intelligent Systems LaboratoryOpen an Excel File• Click the “Import selected sheet to Spreadsheet”• Select the desired Excel sheet where your data is stored• Get variable names from the first row3The University of Iowa Intelligent Systems LaboratoryOpen an Excel File• Change variable typeThe University of Iowa Intelligent Systems LaboratoryOpen an Excel File• Change variable type4The University of Iowa Intelligent Systems LaboratoryClassification and Regression•C&RT• Boosting tree• Neural NetworksThe University of Iowa Intelligent Systems LaboratoryC&RT Classification• Iris data is used as a example data set5The University of Iowa Intelligent Systems LaboratoryC&RT Classification• Click “Data Mining” menu and find the “Interactive Trees”The University of Iowa Intelligent Systems LaboratoryC&RT Classification• View the final tree and understand the results6The University of Iowa Intelligent Systems LaboratoryC&RT---Regression• Use the CPU data set and select the regression analysisDon’t check itThe University of Iowa Intelligent Systems LaboratoryC&RT---Regression• Regression tree structure7The University of Iowa Intelligent Systems LaboratoryC&RT---RegressionPredicted valuesThe University of Iowa Intelligent Systems LaboratoryBoosting tree Classification• In “Data Mining” menu and find the “Boosted Trees”8The University of Iowa Intelligent Systems LaboratoryBoosting tree Classification• See the results and predictor’s importanceThe University of Iowa Intelligent Systems LaboratoryBoosting tree Classification• See the results and predictor’s importance9The University of Iowa Intelligent Systems LaboratoryBoosting tree Regression•CPU data setThe University of Iowa Intelligent Systems LaboratoryBoosting tree Regression• See the results and predictor’s importancePredicted values10The University of Iowa Intelligent Systems LaboratoryBoosting tree Regression• See the results of Observed values vs. Predicted valuesThe University of Iowa Intelligent Systems LaboratoryBoosting tree Regression• See the results and predictor’s importance11The University of Iowa Intelligent Systems LaboratoryNeural Networks Classification• In “Data Mining” menu and find the “Automated Neural Networks”The University of Iowa Intelligent Systems LaboratoryNeural Networks Classification• Choose “Classification”, then select variables12The University of Iowa Intelligent Systems LaboratoryNeural Networks Classification• Statistica will try a set of different neural networks and keep the best onesThe University of Iowa Intelligent Systems LaboratoryNeural Networks Classification• See the classification results13The University of Iowa Intelligent Systems LaboratoryNeural Networks Classification• See the classification results---PredictionsThe University of Iowa Intelligent Systems LaboratoryNeural Networks Classification• See the classification results---Predictions14The University of Iowa Intelligent Systems LaboratoryNeural Networks Classification• See the classification results---Confusion matrixThe University of Iowa Intelligent Systems LaboratoryNeural Networks Regression•CPU data set15The University of Iowa Intelligent Systems LaboratoryNeural Networks Regression• CPU data set, select variablesThe University of Iowa Intelligent Systems LaboratoryNeural Networks Regression• Training and results16The University of Iowa Intelligent Systems LaboratoryNeural Networks Regression• PredictionsThe University of Iowa Intelligent Systems LaboratoryNeural Networks Regression• Some statistics about the