MLD104 · Statistics for Machine Learning

Categories: AI & Automation
Wishlist Share

About Course

The statistics that machine learning depends on: distributions, sampling, uncertainty, hypothesis tests, A/B testing and the bias-variance trade-off, taught with code and business examples.

Build · 6 hours · 4 lessons, a lab, an assessed assignment and a portfolio piece

Who it is for

Analysts and developers who want to reason correctly about data, experiments and model results.

Prerequisites

MLD103 or equivalent experience with pandas and basic charts; high-school algebra.

What you will be able to do

  • Describe data with appropriate distributions and explain sampling variability and the central limit theorem
  • Compute and interpret confidence intervals using formulas and the bootstrap
  • Design and analyze an A/B test, including sample size, p-values and practical significance
  • Explain the bias-variance trade-off and why models are evaluated on data they have not seen

Lessons

  • Lesson 1 · Distributions, sampling and uncertainty
  • Lesson 2 · Confidence intervals, tests and A/B experiments
  • Lesson 3 · Statistics inside machine learning
  • Lesson 4 · Lab and assignment

Lab: Analyzing a pricing A/B test

You will analyze an A/B test from an online store that showed two versions of a checkout page. You will simulate sampling to see the central limit theorem at work, estimate the conversion lift with a confidence interval, run the appropriate test and decide whether the result is worth acting on.

Free tools: Google Colab (free) with NumPy, pandas, SciPy and statsmodels

Assessed assignment

Design and analyze an experiment for a small business decision, such as testing two email subject lines for a dental clinic or two menu layouts for a café, using the provided data or simulated data you justify. State the hypothesis and primary metric before looking at results, compute the required sample size, analyze the results with a confidence interval and a test, and write a one-page decision memo explaining what you recommend and what risks remain.

Criterion Weight What excellent work looks like
Experiment design 25% Hypothesis, primary metric and sample size are defined before analysis and are appropriate for the decision.
Correct analysis 30% Rates, intervals and tests are computed correctly, assumptions are checked and code is reproducible.
Interpretation 25% p-values and intervals are interpreted correctly, with clear distinction between statistical and practical significance.
Decision memo 20% A concise recommendation a manager can act on, with the remaining uncertainty and risks stated plainly.

Portfolio piece

A reproducible A/B test analysis notebook and a one-page decision memo.

Self-paced, 12-month access. Questions are answered in writing through the student panel. A completion record is issued when you finish the course.

Next step

Continue with MLD105 · Supervised Learning: Regression and Classification · 8 hours · $56.00

Continue to the next course

Prefer the whole program? Program in Machine Learning & Applied Data Science · $378.00 (10% less than buying the courses separately)

Show More

What Will You Learn?

  • Describe data with appropriate distributions and explain sampling variability and the central limit theorem
  • Compute and interpret confidence intervals using formulas and the bootstrap
  • Design and analyze an A/B test, including sample size, p-values and practical significance
  • Explain the bias-variance trade-off and why models are evaluated on data they have not seen

Course Content

MLD104 · Statistics for Machine Learning (6 h)
The statistics that machine learning depends on: distributions, sampling, uncertainty, hypothesis tests, A/B testing and the bias-variance trade-off, taught with code and business examples.

  • Lesson 1 · Distributions, sampling and uncertainty
  • Lesson 2 · Confidence intervals, tests and A/B experiments
  • Lesson 3 · Statistics inside machine learning
  • Lesson 4 · Lab and assignment
  • MLD104 Assessment · Statistics for Machine Learning
Scroll to Top