Skip to content
Armand Notue

Statistics · Probability · Coursework

Advanced Probability & Statistics — The foundation under the ML work

Graduate coursework notebooks on descriptive statistics, probability, visualization, hypothesis testing, ANOVA and statistical modeling in Python.

Coursework

At a glance

Status
Coursework
Focus
Eight topic areas · Worked analyses in Python · Hypothesis testing and ANOVA
Stack
  • Python
  • pandas
  • NumPy
  • SciPy
  • Matplotlib
  • seaborn
Problem
Machine-learning results are only as trustworthy as the statistics used to evaluate them.
Approach
A structured collection of Python notebooks organized by topic, each pairing statistical theory with worked analysis on real data.

01

Overview

Notebooks from graduate coursework in the M.S. Data Science program, organized by topic. They are the statistical toolkit behind the evaluation choices in the other projects — significance tests, effect sizes and honest baselines.

02

Problem & motivation

Strong ML evaluation needs strong statistics: knowing when a difference is real, when it is noise, and when a test's assumptions don't hold.

03

My role

  • Coursework — analyses written and run in Python.

04

What's inside

  1. Introduction to statistics and Python (opens in a new tab)

    Python for statistical work, and how misleading data presentations can be detected and corrected.

  2. Descriptive statistics (opens in a new tab)

    Central tendency, spread and outliers on real datasets.

  3. Probability (opens in a new tab)

    Probability theory, bias, and a worked lottery-odds problem.

  4. Data visualization (opens in a new tab)

    Choosing charts for multiple variables, and the ethics of visual presentation.

  5. Statistical modeling (opens in a new tab)

    Regression modeling, its limitations, and correlation versus causation.

  6. Hypothesis testing (opens in a new tab)

    Choosing the appropriate test and checking its assumptions, with SciPy.

  7. ANOVA and linear models (opens in a new tab)

    When to use ANOVA versus regression, and the assumptions behind both.

  8. Exploratory data analysis (opens in a new tab)

    End-to-end EDA, including housing prices and online-retail customers.

05

Technologies used

  • Python
  • pandas
  • NumPy
  • SciPy
  • Matplotlib
  • seaborn

Want the deeper walkthrough?

I'm happy to go through the code, the trade-offs and what I'd do next.

Get in touch