Welcome to Predictive Modeling, Model Fitting, and Regression Analysis. In this course, we will explore different approaches in predictive modeling, and discuss how a model can be either supervised or unsupervised. We will review how a model can be fitted, trained and scored to apply to both historical and future data in an effort to address business objectives. Finally, this course includes a hands-on activity to develop a linear regression model.
Predictive Modeling, Model Fitting, and Regression Analysis
University of California, Irvine via Coursera
Overview
Syllabus
- Predictive Modeling
- Welcome to Module 1, Predictive Modeling. In this module we will begin with a comparison of predictive and descriptive analytics, and discuss what can be learned from both. We will also discuss supervised and unsupervised modeling, two foundational models in analytics and machine learning.
- Data Dimensionality and Classification Analysis
- Welcome to Module 2, Data Dimensionality and Classification Analysis. In this module we will explore how data can be classified and how decision trees can be leveraged as a fast, easy to use a model that is easy to interpret, explain, and visualize.
- Model Fitting
- Welcome to Module 3, Model Fitting. In this module we will explore the concept of model fitting and how creating a generalized model that is able to fit both historical and future data is the ultimate goal. We will also review how a model can be trained or scored to apply to new and unlabeled data.
- Regression Analysis
- Welcome to Module 4, Regression Analysis. In this module we will begin with an explanation of regression analytics, a popular technique used by data science professionals to make predictions. We will also discuss how achieving model fit is not a guarantee that a model can help solve a business problem, and how even a good model can sometimes lead to unactionable outcomes.
Taught by
Julie Pai
