Complete for freeIntermediateFall 2026 · Live online class

Fall 2026 live class

Intermediate ML & Model Evaluation

Intermediate ML & Model Evaluation is a live online class that meets once a week for 1.5 hours over six weeks. Students who have trained a first model learn to tell whether it is actually good: start from a baseline, build leak-free pipelines, choose metrics that fit the problem, validate with cross-validation, compare and tune models, and check how errors are distributed before reporting results. All work happens in free Python notebooks: Google Colab or JupyterLab.

Free completion promise: All class work runs in free tools: Google Colab with a free Google account, or JupyterLab installed locally with no account. The datasets used are public or synthetic, and no GPU, paid cloud service, or paid API is needed.

What's inside

  • 24hand-written lessons
  • 18videos
  • 1step-through animations
  • 80diagrams, charts & live previews
  • 50quiz questions
  • 164code examples

Fall 2026 · Live online class

Class schedule

Meets
Sundays, 5:00–6:30 PM ET
Length
6 weeks · 1.5 hours per class
Starts
First week of November 2026

Fall 2026 registration has closed

Winter 2027 classes start in February 2027. Sign up for Winter 2027 (opens in a new tab)

Week by week

  1. Week 1

    Review the workflow and beat a baseline

    • The supervised learning workflow
    • Train/test splits and baseline models
  2. Week 2

    Prepare data with leak-free pipelines

    • Cleaning data and handling missing values
    • Encoding categories and scaling features
    • Pipelines that prevent data leakage
  3. Week 3

    Choose the right metrics

    • Accuracy, precision, recall, and F1
    • Confusion matrices and decision thresholds
  4. Week 4

    Validate with cross-validation

    • k-fold cross-validation
    • Overfitting, underfitting, and learning curves
  5. Week 5

    Compare and tune models

    • Comparing decision trees, random forests, and linear models
    • Hyperparameter tuning with grid and random search
  6. Week 6

    Handle imbalance and report honestly

    • Handling class imbalance
    • Error analysis across slices of the data
    • Communicating results and limitations

Before the first class, finish the free setup module. After the last class, you'll complete and present the capstone project.

Free tools you'll use in class

After the class: keep going at your own pace with Intermediate Machine Learning and Model Evaluation.

Before you begin

Requirements & setup for Intermediate ML & Model Evaluation

Start with the complete free pathway below. Hardware, subscriptions, paid APIs, and cloud credits are never required.

Start here · free path

Your complete $0 setup

All class work runs in free tools: Google Colab with a free Google account, or JupyterLab installed locally with no account. The datasets used are public or synthetic, and no GPU, paid cloud service, or paid API is needed.

$0 required

Setup checklist

0 of 5 complete

This checklist is a temporary setup aid and resets when the page reloads. Saved course progress begins in the setup lessons below.

  1. Everything required for this course is free. Paid software, APIs, cloud credits, GPUs, and physical equipment are optional.

    About 2 min
  2. Go to colab.research.google.com and sign in with a Google account.

  3. Choose File > New notebook.

    About 5 min
  4. Run import sklearn, pandas; print(sklearn.__version__) in the first cell.

    About 5 min
  5. Google Colab: Run a cell that prints the scikit-learn version.

    About 5 min

Software

1 item

Google Colab

Free Python notebooks in the browser with pandas, scikit-learn, and matplotlib already installed. Needs a free Google account; JupyterLab on your own computer works too.

Course need
Google Colab
Free

Free Python notebooks in the browser with pandas, scikit-learn, and matplotlib already installed. Needs a free Google account; JupyterLab on your own computer works too. Every lesson is readable text with copyable code, keyboard-friendly quizzes and step-throughs, and videos with YouTube captions.

  1. Go to colab.research.google.com and sign in with a Google account.
  2. Choose File > New notebook.
  3. Run import sklearn, pandas; print(sklearn.__version__) in the first cell.

Hardware

Nothing required

No physical hardware is needed for this course.

Accounts

Nothing required

No external account is needed for this course.

What you will be able to do

Learning outcomes

  • Compare models against a baseline using a clean train/test workflow.
  • Build scikit-learn pipelines that prepare data without leaking information.
  • Choose, compute, and explain metrics, thresholds, and cross-validation results.
  • Tune and compare models, handle imbalance, and report errors and limitations honestly.

Build real work

Assignments & capstone

Free tools accepted

Baseline and pipeline notebook

Build a notebook that compares a baseline with a real model, using a leak-free preprocessing pipeline on a public dataset.

All class work runs in free tools: Google Colab with a free Google account, or JupyterLab installed locally with no account. The datasets used are public or synthetic, and no GPU, paid cloud service, or paid API is needed. Submit your project file, or screenshots with descriptions, or a short demo.

Metric and validation report

Evaluate a classifier with a confusion matrix, precision, recall, F1, a chosen threshold, and cross-validated scores, and explain what the numbers mean.

All class work runs in free tools: Google Colab with a free Google account, or JupyterLab installed locally with no account. The datasets used are public or synthetic, and no GPU, paid cloud service, or paid API is needed. Submit your project file, or screenshots with descriptions, or a short demo.

Final assessment

End-to-end model evaluation report

Choose a public dataset, build and tune at least two models in a leak-free pipeline, evaluate them with cross-validation and well-chosen metrics, analyze errors across slices, and present a recommendation with its limitations. Deliverables: A notebook that runs from top to bottom, A baseline and at least two tuned models in pipelines, Cross-validated metrics with a justified threshold, Slice-based error analysis, and A short written or spoken recommendation with limitations.

Prerequisites & pacing

Skills assumed

  • Python basics: variables, loops, functions, and lists
  • Some experience with pandas or training a first scikit-learn model

Recommended prior courses

These are advisory, not locks. If you already have the skills listed above, enroll directly.

Choose your pace

  • self paced: 6 weeks at 1.5 hours per week
  • weekly: 6 weeks at 1.5 hours per week

Course created by Zafir Shamsi.