Machine Learning Series Roadmap

August 4, 2025


Hi there 👋🏼 Welcome to the ML/AI Series by Rohan Tinna. In this series, I'm going to cover the foundational concepts in artificial intelligence, building up from the basics to the more advanced topics.

In order to delve into these concepts, you'll need to be familiar with topics in statistics & probability theory, linear algebra, and calculus. I have listed the topics we'll need in understanding the material covered in this series below. I'll also be covering these topics as needed but those videos will be added to the respective playlist. For example, before introducing linear regression, I will cover data distributions in a seperate video that will be added to the Statistics & Probability Theory Playlist and then make the linear regression video. I'll be using Python for this series. If you want to learn python, see Python resources.

Below is a tentative plan I have curated. Checkout the additional resources.

Playlists

Playlist
Linear Algebra for ML
Calculus for ML
Statistics and Probability Theory
Machine Learning & AI

Tentative Plan

* Subject to change

Building Blocks

  1. Statistics & Probability Theory -> Common data distributions, Linear Regression,
  2. Linear Algebra
  3. Calculus
TopicPlaylistDataVideo
1. Introduction to the courseML/AITBA
2. Data distributions – Visualizing dataStatistics & Probability TheoryHeights & WeightsTBA
3.Linear Regression – The Least Squares MethodStatistics & Probability Theory; ML/AITBA
4. Logistic RegressionStatistics & Probability Theory; ML/AITBA
5. K-nearest neighbours model for classificationML/AITBA

Prediction/Learning Models

  • Linear Regression
  • Logistic Regression
  • Cluster Analysis
  • State Vector Machines
  • k-Nearest Neighbours
  • Simple Neural Network

Deep Learning

  • Brief History of Deep Learning
  • Multi-Layer Perceptron
  • Intro to Generative AI Applications

Introduction

Machine Learning (ML) is a relatively recent field where we use computers to learn from data and make predictions. This abstract goal can be applied to a wide range of situations to automate tasks or optimize processes. My tone may be a bit abstract now, but I'll try to give real-world examples as we go along.

ML is a fascinating field for me because it lies at the intersection of computer science, statistics, and domain-specific knowledge. We use trivial statistical concepts to build models so complex that they seem as good as humans, if not better, at certain tasks, which still amazes me.

I'll also be sharing anecdotes from the history of the AI industry which will add entertainment and charm to the series.

Y=f(x)+ϵ\begin{aligned} Y = f(x) + \epsilon \end{aligned}
flowchart TD 0[Reasons we estimate $$f$$] --> 1[Prediction] 0 --> 2[Inference] --> 2.1@{ shape: text, label: "Inference description" } 1.1[Parametric] 1.2[Non-Parametric] 1 --> 1.1 & 1.2