Python Programme
Data Modeling with Python
Build and diagnose statistical models with statsmodels, from regression through generalized linear models to the basics of time series.
Course Overview
Machine learning models are often built to predict; statistical models are built to explain. This course teaches you the second half of that pair: how to build models in Python whose coefficients, diagnostics, and assumptions you can defend, not just models that score well on a test set.
You'll work with statsmodels to fit and diagnose linear and generalized linear models, compare and select between candidate models, and get a practical introduction to time series, understanding trend, seasonality, and simple forecasting.
This is the final course in our upcoming Python track, building directly on Machine Learning with Python. If you haven't taken that course, start with Python for Data Analysis first.
Course at a Glance
What You'll Learn
A step-by-step path from your first statistical model to defensible, correctly diagnosed results.
Statistical Modeling Foundations
Understand the difference between predictive and explanatory modeling, and get set up with statsmodels.
Linear Regression & Diagnostics
Fit OLS regression models, interpret coefficients, and check assumptions through residual analysis.
Logistic & Generalized Linear Models
Model binary and count outcomes with logistic and Poisson regression, and interpret coefficients correctly.
Model Comparison & Selection
Compare candidate models using AIC and BIC, and apply regularisation with Ridge and Lasso regression.
Time Series Modeling Basics
Identify trend and seasonality in time-ordered data and build simple forecasting models.
Modeling Capstone
Build, diagnose, and present a statistical model on a real research or business dataset from start to finish.
Who This Course Is For
Built for people who need to move beyond prediction to explanation, and defend their results.
Prerequisites
Completion of Machine Learning with Python, or an equivalent, working understanding of Python, pandas, and core machine learning concepts, is required before starting this course.
Want to be first in line?
This course is in development as part of our upcoming Python track. Register your interest and we'll notify you as soon as enrolment opens.