This is an advanced training designed for people with intermediate to experienced data analysis and statistics knowledge. Machine learning and predictive modeling utilizes statistics and mathematics in order to predict future outcomes. It is the core technology behind key analytic processes and models such as database marketing, customer loyalty and retention, customer churn and win back, credit risk analysis and management etc.
- Multiple linear regression, logistic regression, decision tree, neural network, deep learning.
- Model diagnostics and residual analysis
- Model building process: performance definition, data sampling, data cleansing, variable transformation/imputation, missing values /outliers, WOE technique, variable selection procedure.
- Model validation: cross validation, decile analysis, lift curve, odds chart, KS, GINI , ROC, AUC.
- Model implementation and strategies
In contrast to university courses, this training pays close attention to the practical uses and approaches of statistical modeling in real industries. Instead, we focus on practical modeling methods, real work practice, business insights, interpretation and application. It includes 3 hands-on projects which will allow you to grasp these advanced concepts in no time.