Schulung - IBM 0A079G - Introduction to Machine Learning Models Using IBM SPSS Modeler (V18.2)

Course Outline

Introduction to machine learning modelsTaxonomy of machine learning modelsIdentify measurement levelsTaxonomy of supervised modelsBuild and apply models in IBM SPSS ModelerSupervised models: Decision trees - CHAIDCHAID basics for categorical targetsInclude categorical and continuous predictorsCHAID basics for continuous targetsTreatment of missing valuesSupervised models: Decision trees - C&R TreeC&R Tree basics for categorical targetsInclude categorical and continuous predictorsC&R Tree basics for continuous targetsTreatment of missing valuesEvaluation measures for supervised modelsEvaluation measures for categorical targetsEvaluation measures for continuous targetsSupervised models: Statistical models for continuous targets - Linear regressionLinear regression basicsInclude categorical predictorsTreatment of missing valuesSupervised models: Statistical models for categorical targets - Logistic regressionLogistic regression basicsInclude categorical predictorsTreatment of missing valuesSupervised models: Black box models - Neural networksNeural network basicsInclude categorical and continuous predictorsTreatment of missing valuesSupervised models: Black box models - Ensemble modelsEnsemble models basicsImprove accuracy and generalizability by boosting and baggingEnsemble the best modelsUnsupervised models: K-Means and KohonenK-Means basicsInclude categorical inputs in K-MeansTreatment of missing values in K-MeansKohonen networks basicsTreatment of missing values in KohonenUnsupervised models: TwoStep and Anomaly detectionTwoStep basicsTwoStep assumptionsFind the best segmentation model automaticallyAnomaly detection basicsTreatment of missing valuesAssociation models: AprioriApriori basicsEvaluation measuresTreatment of missing valuesAssociation models: Sequence detectionSequence detection basicsTreatment of missing valuesPreparing data for modelingExamine the quality of the dataSelect important predictorsBalance the data

Objective

Introduction to machine learning models Taxonomy of machine learning models Identify measurement levels Taxonomy of supervised models Build and apply models in IBM SPSS Modeler 

Supervised models: Decision trees - CHAID CHAID basics for categorical targets Include categorical and continuous predictors CHAID basics for continuous targets Treatment of missing values 

Supervised models: Decision trees - C&R Tree 

C&R Tree basics for categorical targets Include categorical and continuous predictors C&R Tree basics for continuous targets Treatment of missing values 

Evaluation measures for supervised models Evaluation measures for categorical targets Evaluation measures for continuous targets 

Supervised models: Statistical models for continuous targets - Linear regression Linear regression basics Include categorical predictors Treatment of missing values 

Supervised models: Statistical models for categorical targets - Logistic regression Logistic regression basics Include categorical predictors Treatment of missing values

Association models: Sequence detection Sequence detection basics Treatment of missing values

Supervised models: Black box models - Neural networks Neural network basics Include categorical and continuous predictors Treatment of missing values  

Supervised models: Black box models - Ensemble models Ensemble models basics Improve accuracy and generalizability by boosting and bagging Ensemble the best models  

Unsupervised models: K-Means and Kohonen K-Means basics Include categorical inputs in K-Means Treatment of missing values in K-Means Kohonen networks basics Treatment of missing values in Kohonen  

Unsupervised models: TwoStep and Anomaly detection TwoStep basics TwoStep assumptions Find the best segmentation model automatically Anomaly detection basics Treatment of missing values  

Association models: Apriori Apriori basics Evaluation measures Treatment of missing values

Preparing data for modeling Examine the quality of the data Select important predictors Balance the data