Predictive Modeling and Text Mining

Predictive analytics is about using data and statistical algorithms to predict what might happen next given the current process and environment.

In this module, you will learn about some of the core techniques used in building predictive models, including how to address overfitting, select the best predictive model, and use multiple linear regression and logistic regression. You will also see how to fit other types of predictive models, including penalized regression, decision trees and neural networks. Finally, you will learn how to extract information and meaning from unstructured text data, such as survey response data.

Estimated time to complete this module: 3 to 4 hours

Enroll now

Specific topics covered in this module include:

Essentials of Predictive Modeling

Introduction to Predictive Modeling
Overfitting and Model Validation
Assessing Model Performance: Prediction Models
Assessing Model Performance: Classification Models
Receiver-Operating Characteristic (ROC) Curves

Decision Trees

Introduction to Decision Trees
Classification Trees
Regression Trees
Decision Trees with Validation
Random (Bootstrap) Forests

Neural Networks

What is a Neural Network?
Interpreting Neural Networks
Predictive Modeling with Neural Networks

Generalized Regression

Introduction to Generalized Regression
Fitting Models Using Maximum Likelihood
Introduction to Penalized Regression

Model Comparison and Selection

Comparing Predictive Models

Introduction to Text Mining

Introduction to Text Mining
Processing Text Data
Curating the Term List
Visualizing and Exploring Text Data
Analyzing (Mining) Text Data