← Free Learning Library

Machine Learning

Machine Learning Crash Course: Core Concepts

A short video series from Google for Developers covering the fundamentals of machine learning, from gradient descent to neural networks and fairness.

Learn from: the original creator. Original content published on YouTube. DigitalSkillX organizes these public resources into a structured learning path. DigitalSkillX does not claim ownership or partnership.

Lessons
10
Time
About 21 min
Level
beginner

Now playing: Machine Learning Crash Course: Intro & What's New

About this path

This learning path follows the Machine Learning Crash Course video series from Google for Developers. It walks through core machine learning ideas in order: training with gradient descent, logistic regression, classification, generalization, neural network basics and backpropagation, embeddings, and fairness. The videos are brief and example-driven, making the path useful as a first pass or a refresher. Most lessons are a few minutes long, with one very short bonus clip about the ReLU activation function.

What you will learn

  • Explain how gradient descent trains a model by optimizing parameter values
  • Describe how logistic regression predicts probabilities
  • Set a classification threshold and interpret classification results
  • Recognize overfitting and understand what generalization means
  • Identify the building blocks of neural networks: nodes, hidden layers, and activation functions
  • Understand how backpropagation optimizes neural network parameters
  • Explain what embeddings are and how they encode relationships between values
  • Describe basic principles for auditing ML models for fairness

Who this is for

Beginners who want a structured free introduction before deeper practice.

Curriculum

Section 1: Getting Started

  1. Lesson 1: Machine Learning Crash Course: Intro & What's New

    This intro video announces a refreshed version of Machine Learning Crash Course. It notes that millions of people have used the course since 2018 and that the update covers recent advances in AI. Use it as a starting point before moving into the technical lessons.

    Open on YouTube

Section 2: Training and Prediction

  1. Lesson 2: Machine Learning Crash Course: Gradient Descent

    Gradient descent is an algorithm for training machine learning models by iteratively optimizing their parameter values. The video shows how to use gradient descent to train a linear regression model that predicts travel times. It is a foundational lesson for understanding how models learn from data.

    Open on YouTube
  2. Lesson 3: Machine Learning Crash Course: Logistic Regression

    Logistic regression is a technique for predicting a probability. The video uses the example of predicting whether an email message is spam. It connects the method to the broader Machine Learning Crash Course material.

    Open on YouTube
  3. Lesson 4: Machine Learning Crash Course: Classification

    Classification is a machine learning technique for predicting a class or category, such as spam or not spam. The video explains how to set a classification threshold to convert model output into a class prediction. It builds on the probability output from logistic regression.

    Open on YouTube
  4. Lesson 5: Machine Learning Crash Course: Generalization

    Generalization is a model's ability to make good predictions on data it has not seen before. The video walks through a scenario where a model fails to generalize and shows how to address the problem. It highlights why overall training performance is not the whole story.

    Open on YouTube

Section 3: Neural Networks

  1. Lesson 6: Machine Learning Crash Course: Neural Networks Intro

    Neural networks are models that can automatically learn nonlinear relationships in data. This video introduces the building blocks of neural network architectures: nodes, hidden layers, and activation functions. It sets up the later lesson on backpropagation.

    Open on YouTube
  2. Lesson 7: Machine Learning Crash Course: Neural Networks Backprop

    Backpropagation is a common algorithm for optimizing parameter values in a neural network. The video explains how it works using a classroom analogy of students collaborating. It follows the neural networks intro lesson.

    Open on YouTube
  3. Lesson 8: Machine Learning Crash Course: Embeddings

    An embedding translates large feature vectors into a lower-dimensional space that encodes meaningful relationships between values. The video shows how to train an embedding that represents relationships between words. It is a useful introduction to how embeddings capture similarity.

    Open on YouTube

Section 4: Responsible ML

  1. Lesson 9: Machine Learning Crash Course: Fairness

    This lesson covers principles and best practices for auditing ML models for fairness. It includes strategies for identifying and mitigating biases in data. It also notes that strong overall performance can hide problems for specific groups.

    Open on YouTube
  2. Lesson 10: Machine Learning Crash Course: The Optimistic ReLU Function

    A very short clip that jokes about why the ReLU activation function is always in a good mood. It points viewers to the Machine Learning Crash Course for more on ReLU, activation functions, and neural networks. Treat it as a light bonus rather than a full lesson.

    Open on YouTube

Your progress

0 of 10 lessons completed (0%).

Complete this learning path to become eligible for your DigitalSkillX certificate.

Related learning