Regression with R

TRAINING COURSE

Details

Performing regressions is a powerful method to help you understand the relationships between your variables, which makes it a must for anyone working with data. Thankfully, we can use the Python language to make this important task much easier and more efficient.

In this course you will learn to: 

  • Perform regressions using Python
  • Analyse the strength of correlations
  • Set conditional expectations
  • Analyse the regression line
  • Employ linear models in real-life scenarios 
  • Understand counfounding
  • Perform Lease Squares Estimates
  • Understand association vs causation

Delivery Methods

Delivery Method Duration
Classroom
4 Days Get a Quote
Live Virtual Training
4 Days Get a Quote

Discounts Available

Brochure:

Download Brochure

Information may change without notice.

Audience

  • Data scientists
  • Programmers
  • Business analysts 
  • Engineers
  • Scientists

Pre-Requisites

Leading Training's Introduction to R Programming or equivalent knowledge

Course Outline / Curriculum

Regression 

  • Case study: is height hereditary?
  • The correlation coefficient
    • Sample correlation is a random variable
    • Correlation is not always a useful summary
  • Conditional expectations
  • The regression line
    • Regression improves precision
    • Bivariate normal distribution (advanced)
    • Variance explained
    • Warning: there are two regression lines
  • Exercises

Linear models 

  • Case study: Moneyball
    • Sabermetics
    • Baseball basics
    • No awards for BB
    • Base on balls or stolen bases?
    • Regression applied to baseball statistics
  • Confounding
    • Understanding confounding through stratification
    • Multivariate regression
  • Least squares estimates
    • Interpreting linear models
    • Least Squares Estimates (LSE)
    • The lm function
    • LSE are random variables
    • Predicted values are random variables
  • Exercises
  • Linear regression in the tidyverse
    • The broom package
  • Exercises
  • Case study: Moneyball (continued)
    • Adding salary and position information
    • Picking nine players
  • The regression fallacy
  • Measurement error models
  • Exercises

Association is not causation 

  • Spurious correlation
  • Outliers
  • Reversing cause and effect
  • Confounders
    • Example: UC Berkeley admissions
    • Confounding explained graphically
    • Average after stratifying
  •  Simpson’s paradox
  • Exercises

Schedule Dates and Booking

There are currently no scheduled dates.

Add me to the waiting list

Submit Enquiry

Name
Email
Telephone
Query