Artificial Intelligence & Machine Learning from scratch

Welcome to the very first lesson of the "AI & ML course from scratch" by Dr. Long Nguyen!

In this lesson, I will show you some the motivation and the definition of AI as well as some of its examples.

This is a very important lesson!

AI is a very big domain with so many fields that are sometimes confusing. For example today people are talking a lot about AI, Machine Learning, Deep Learning or Reinforcement Learning ... but I know that there are still many people misunderstanding the relationship among these concepts.

By this lesson, you will understand the two different ways to classify AI, and explore its different branches in order to get a correct overview of the domain.

As a matter of fact, Machine Learning may be the most important sub-field of AI today, with so many real-life applications.

So in this lesson, I will introduce to you the motivation and the definitions of Machine Learning.

In this lesson, I am going to present to you the three most important types of Machine Learning, they are: Supervised Learning, Unsupervised Learning, and Reinforcement Learning.

Linear Regression is a very fundamental tool in Supervised Learning, with so many applications, notably in predictive analysis.

In this lesson, I am going to show you the basic concepts of Linear Regression through a use case.

In this lesson, I will present to you one of the most popular methods to optimize the cost function in Linear Regression, it is called Gradient Descent.

In this lesson, you will learn about Multiple Linear Regression, that is, linear regression with more than one variables.

Also, I will give you an interesting point of view about Polynomial Regression.

Logistic Regression is a very basic method for binary classification.

This lesson will show you the concept of decision boundary, which is needed to understand Logistic Regression.

In this lesson, I will show you how we define the sigmoïde activation function, the cost function, and how to optimize the cost function in Logistic Regression.

In the previous parts, you already learned about Binary classification in which there are only two output classes.

This lesson will present to you the One-versus-All method which is used for Multinomial classification (more than two output classes).

Welcome to the lesson about one of the hottest technology terms today: Artificial Neural Network (ANN)!

In this first part, I am going to present to you the motivation and the definition of ANN, then we will design a simple neuron together.

In this lesson, I will show you simple and complex Artificial Neural Networks, with a step-by-step calculation example.

Also, you will know what is Deep Learning, a very hot concept today.

In this lesson, I am going to show you some important notations that we will use through the lesson.

Then, we will discuss about Multinomial classification using Artificial Neural Network, and define the cost function for the problem.

In this lesson, you will learned about Backpropagation algorithm, a very effective method to calculate partial derivatives of the cost function in Artificial Neural Network, so that we can use Gradient Descent to minimize the cost function in order to find the optimal connection weights.

In this lesson, I will introduce to you the most typical problem in Unsupervised Learning, it is Clustering analysis which has so many important applications in real life.

You will learn about the most basic approach in clustering analysis: the K-means algorithm.

Indeed, running K-means with different random initializations may give different clustering results because K-means algorithm sometimes finds local minima rather than the global minimum. So how do people often do in practice?

Moreover, how can we choose the best number of clusters K?

These questions will be answered in this lesson.

Welcome to the world of beautiful logic!

By this lesson, you will find that Fuzzy Logic is very friendly and close to our real life. It is a classical, but a very nice approach in Artificial Intelligence and still has many important applications today.

Fuzzy control system is a very important aspect in Fuzzy Logic.

Fuzzy control allows us to easily design an intelligent system that can reuse expert's experiences.

In this lesson, you will learn about the three most important stages in a Fuzzy system: Fuzzification, Inferences, and Defuzzification.

This lesson will explain you what is the Inference stage in a Fuzzy control system through intuitive examples.

In this lesson, you will learn about Defuzzification, the last stage in a Fuzzy control system.

Welcome to a biological class!

In this lesson, you will know how the Biological evolution or the Evolution theory (of Darwin) inspired the creation of a very important branch in AI: Evolutionary Computation.

This lesson will introduce to you a very famous algorithm in AI: Genetic Algorithm which is inspired by the evolution theory of Darwin.

In this lesson, we start going into the detail of Genetic Algorithm with chromosome representation, fitness function, and initialization.

This lesson will show you how we use selection, crossover, and mutation to repeatedly create new generations from the initialized population.