BUEX-C531 Predictive Analytics/Data Mining
- 12 Weeks
This course on data mining andpredictive analytics provides students with both the conceptual underpinningsof a broad variety of data mining models as well as experience with analyzingreal data sets.
Topics include: data preparation,cleaning, and exploratory analysis using data visualization and descriptivestatistics; applications of multiple regression for numeric prediction;building predictive models using logistic regression, k-nearest neighbors,naive Bayes, classification and regression trees, neural nets, discriminantanalysis, advanced predictive techniques based on ensembles of predictions suchas bagging and boosting, and selected time series forecasting methods; findingpatterns in data using unsupervised models including principle components andcluster analysis; evaluating the performance of predictive models usingtraining, validation, and testing data subsets as well as k-foldcross-validation; evaluating the performance of predictive and classificationmodels using Receiver Operating Characteristic (ROC), lift charts andstatistics on confusion matrices.
The class sessions provideoverview of the theory behind each model as well as demonstrations using Excel,Power BI, JMP Pro from SAS, and other tools.