SAS Certified Statistical Business Analyst Using SAS 9
This course is designed for SAS professionals who use SAS/STAT software to conduct and interpret complex statistical data analysis. It covers analysis of variance, linear and logistic regression, preparing inputs for predictive models, and measuring model performance.Duration: 40-45hrs
Course Content:
Statistics 1: Introduction to ANOVA, Regression, and Logistic Regression
Course Overview and Review of Concepts
- Descriptive statistics.
- Inferential statistics.
- Examining data distributions.
- Obtaining and interpreting sample statistics using the UNIVARIATE procedure.
- Examining data distributions graphically in the UNIVARIATE and FREQ procedures.
- Constructing confidence intervals.
- Performing simple tests of hypothesis.
- Performing tests of differences between two group means using PROC TTEST.
- Performing one-way ANOVA with the GLM procedure.
- Performing post-hoc multiple comparisons tests in PROC GLM.
- Producing correlations with the CORR procedure.
- Fitting a simple linear regression model with the REG procedure.
- Performing two-way ANOVA with and without interactions.
- Understanding the concepts of multiple regression.
- Automated model selection techniques in PROC GLMSELECT to choose from among several candidate models.
- Interpreting and comparison of selected models.
- Examining residuals.
- Investigating influential observations.
- Assessing collinearit.
- Understanding the concepts of predictive modeling.
- Understanding the importance of data partitioning.
- Understanding the concepts of scoring.
- Obtaining predictions (scoring) for new data using PROC GLMSELECT and PROC PLM.
- Producing frequency tables with the FREQ procedure.
- Examining tests for general and linear association using the FREQ procedure.
- Understanding exact tests.
- Understanding the concepts of logistic regression.
- Fitting univariate and multivariate logistic regression models using the LOGISTIC procedure.
- Using automated model selection techniques in PROC LOGISTIC including interaction terms.
- Obtaining predictions (scoring) for new data using PROC PLM.
Predictive Modeling Using Logistic Regression
- business applications
- analytical challenges
- parameter estimation
- adjustments for oversampling
- missing values
- categorical inputs
- variable clustering
- variable screening
- subset selection
- ROC curves and Lift charts
- optimal cutoffs
- K-S statistic
- c statistic
- profit
- evaluating a series of models