Data analyst
Add predictive models to existing analysis.
Harvard / Harvard Online / HarvardX
Build AI systemsMachine Learning and AI with Python
Apply tree-based machine learning to datasets and judge model performance.
Curriculum
- Decision trees; bagging and random forests
- Gradient boosting
- Bias, overfitting and evaluation
Before you start
Python and statistics; familiarity with bootstrapping, multilogistic regression, hyperparameters and missing data.
Who it suits
Our assessment, based on the syllabus.
Operations or finance analyst
Evaluate operational predictions using data.
Developer working with tabular data
Apply conventional ML without a long degree.
Expected outcomes
- Train and compare introductory predictive models.
- Recognise biased data and underfitting/overfitting.
- Evaluate predictions against held-out results.
Example application
Compare predictive models for identifying customer accounts that may need follow-up.
Limits
- Not suitable as a first course without Python and statistics.
- Predictive ML focus; not agents or coding assistants.
How current is it?
A recent curriculum revision date could not be confirmed.
When to choose it
Consider this if you know Python and statistics and want to add predictive models to your analysis.
Sources
- Official course page ↗harvardonline.harvard.edu
Checked 20 October 2026. Confirm the current offer on the provider’s page before paying.