Identify the most important predictive variables in a model
Quantify qualitative variables and incorporate them in a predictive model
Build predictive models to anticipate trends and demand
|Prerequisites||Intro to R|
|Instruction||3 hours and 30 minutes|
|Practice||20 to 25 hours|
Syllabus: Regression and Time Series Analysis
This course is designed for students who have taken Data Society’s Introduction to R and Visualization course or have a good knowledge of R programming. This 3 ½ hour course teaches students how to apply advanced regression and time series models to accurately predict business trends and demand.
By the end of this course, students will be able to:
1. Identify the most important predictive variables in a model
2. Quantify qualitative variables and incorporate them in a predictive model
3. Build predictive models to anticipate trends and demand
- Concept reviews: these are comprised of short five question quizzes that cover the most important concepts and ideas in each lesson. They encourage holistic understanding and are multi-faceted ques=on types (i.e. drag and drop, fill-in-the-blanks, matching, etc).
- Exercises: these are additional videos that cover the coding functions in the instructional video in more depth. They are project-based and include coding templates for students to strengthen their skills outside of the course.
- Accompanying PDFs to use as reference materials
- R code templates from the instructional videos and exercises
- Data sets used in the instructional videos and exercises
1. Setting up linear regression (34 min) – Free trial!
Overview of data science
What is regression?
Building linear regression
Assessing your model’s accuracy
2. Measuring model errors (38 min)
Calculating variance and standard deviation
Calculating covariance and correlation
Testing a model’s significance
3. Modeling multiple variables (31 min)
Building a multivariate model
Plotting a multivariate regression
Datafying categorical variables
4. Adjusting your model (38 min)
Validating your model
Testing for heteroscedasticity
Identifying important variables
Building nonlinear regression models
5. Adding seasonality to your model (36 min)
Minimizing seasonality errors
Refining your model (30 min)
Calculating seasonality components
Predicting customer demand
Using the LOESS method
Additional considerations and tips
Total instructional time: 3 hrs, 27 min
Dr. Kaska Adoteye
Kaska Adoteye has a PhD in Applied Mathematics from NC State University. He is passionate about using math and statistics to solve real world problems with a maximum impact. He has formulated, tested, and coded models to profitably trade stocks for an investment firm, as well as to create predictive models to examine the effects of various toxicants on populations of biological organisms. He is currently developing an artificial intelligence engine to help secure cyber networks.