Quantitative & Verbal
Reasoning for AI/ML
An 8-Week Bootcamp — Detailed Lesson Plans & Exercises
Course Overview
Build the Foundations That Matter
Every concept taught through the lens of machine learning — not abstract mathematics.
This 8-week bootcamp builds the quantitative and verbal reasoning foundations that every AI/ML practitioner needs. Students will leave with the ability to reason about models, read technical papers, interpret results, and communicate AI concepts clearly.
Session Structure: 30 min concept lesson + 30 min guided exercises + 30 min practice & peer discussion
Target Audience: Aspiring AI/ML practitioners who can follow tutorials but lack the reasoning foundations to build, debug, and communicate about models independently.
Course Schedule
8 Weeks. 15 Sessions.
| Week | Theme | Sessions |
|---|---|---|
| Week 1 | Numbers & Proportions | Session 1: Percentages, Ratios & Rates of Change Session 2: Functions, Inputs & Outputs |
| Week 2 | Describing Data | Session 3: Mean, Median, Mode & Spread Session 4: Distributions & What Data Looks Like |
| Week 3 | Data in Context | Session 5: Putting It All Together — Reading a Dataset Session 6: Thinking in Probabilities |
| Week 4 | Probabilistic Reasoning | Session 7: Bayesian Thinking & Updating Beliefs Session 8: Vectors & Matrices — No Fear |
| Week 5 | How Models Learn | Session 9: Loss Functions & Optimization Intuition Session 10: Evaluation Metrics — Beyond Accuracy |
| Week 6 | Reading & Interpreting | Session 11: Decoding ML Papers & Documentation Session 12: Logical Reasoning — If/Then, Fallacies |
| Week 7 | Communicating ML | Session 13: Technical Comm for Non-Technical Audiences Session 14: Technical Writing — Experiment Reports |
| Week 8 | Capstone | Session 15: Critical Paper Review (Capstone) |
Syllabus
Choose a Week
Each week builds on the last. Click to view full lesson plans, exercises, and discussion prompts.
Numbers & Proportions
Session 1: Percentages, Ratios & Rates of Change
Session 2: Functions, Inputs & Outputs
Describing Data
Session 3: Mean, Median, Mode & Spread
Session 4: Distributions & What Data Looks Like
Data in Context
Session 5: Putting It All Together — Reading a Dataset
Session 6: Thinking in Probabilities
Probabilistic Reasoning
Session 7: Bayesian Thinking & Updating Beliefs
Session 8: Vectors & Matrices — No Fear
How Models Learn
Session 9: Loss Functions & Optimization Intuition
Session 10: Evaluation Metrics — Beyond Accuracy
Reading & Interpreting
Session 11: Decoding ML Papers & Documentation
Session 12: Logical Reasoning — If/Then, Fallacies
Communicating ML
Session 13: Technical Comm for Non-Technical Audiences
Session 14: Technical Writing — Experiment Reports
Capstone
Session 15: Critical Paper Review (Capstone)
View lessonsAssessment
Assessment Rubric
Students are assessed on demonstrated reasoning, not rote calculation.
| Dimension | Developing | Proficient | Advanced |
|---|---|---|---|
| Quantitative Reasoning | Can perform calculations but struggles to interpret results | Interprets results in context and identifies when numbers are misleading | Independently questions methodology and identifies subtle quantitative issues |
| Probabilistic Thinking | Understands basic probability but confuses conditional relationships | Correctly applies Bayes' reasoning and identifies base rate issues | Naturally frames ML problems probabilistically and catches fallacies |
| Technical Reading | Extracts main claim but misses nuances and hedging | Identifies claims, evidence, limitations, and hedging language | Critically evaluates methodology and identifies unstated assumptions |
| Technical Communication | Explains concepts but uses excessive jargon or vague language | Adapts explanations to audience and structures writing clearly | Communicates complex ideas simply with honest uncertainty framing |
Interactive Tools
Practice & Explore
Hands-on tools that make abstract concepts concrete.
Bayes' Theorem
Enter prior, likelihood, and evidence. Watch the posterior update step by step.
Open tool BuilderConfusion Matrix
Enter TP, TN, FP, FN. See accuracy, precision, recall, and F1 calculated instantly.
Open tool VisualizerTraining Loss
See a realistic training curve. Identify where convergence happens and where overfitting begins.
Open tool ReferenceGlossary
Every term from the course, defined in plain English, with links back to where it is taught.
OpenAppendix
Instructor Notes
Session 5 — Integration Day
Use any publicly available tabular dataset with class imbalance. The UCI Machine Learning Repository has several suitable options, such as credit card fraud datasets. Print summary statistics — students don't need the raw data or code access.
Session 15 — Capstone
Select a short paper or blog post (under 5 pages) with clear experimental methodology. Good sources include Distill.pub articles, short workshop papers, or ML blog posts from major labs that include evaluation tables. Avoid papers that require advanced math to understand the core claim.
Pacing Guidance
With two sessions per week, students have multiple days between sessions to let concepts settle. Encourage students to spend 30–45 minutes between sessions reviewing their exercise solutions and attempting any problems they didn't finish during class.
Checkpoints
Weeks 3 and 5 are natural checkpoints. Session 5 (Integration Day) tests whether Weeks 1–2 landed. Session 10 (Evaluation Metrics) closes the quantitative arc — students should feel confident with numbers before the verbal reasoning shift in Week 6.
Transition to Verbal Reasoning
The transition from quantitative to verbal reasoning in Week 6 can feel abrupt. Bridge it by emphasizing during Sessions 9–10 that "the numbers mean nothing if you can't communicate them." This frames verbal reasoning as the natural next step, not a separate topic.