Machine Learning and Deep Learning Using TensorFlow

Machine Learning and Deep Learning Using TensorFlow


Machine Learning, Neural Networks, Deep Learning, Deep Neural Networks (DNN), and Convolution Neural Networks (CNN)

What you'll learn
In depth understanding of Machine Learning.
In depth understanding of the Neural Network.
Detailed and step by step theoretical derivation and explanation of a majority of the topics to ensure clear understanding of the subject.
You will learn Linear Regression, Logistic Regression, Neural Network, Deep Neural Network (DNN), Convolution Neural Network etc.
Multiple hands-on projects using Tensorflow 2 and Python to expose you to some of the highly advanced topics of Tensorflow 2
Hands-on projects are selected to make you familiar with some of the expertise that may be very useful should you need to run a very long analysis in future.

Requirements
For Theory Section: Some knowledge in algebra and calculus.
For Hands-On Section: Gmail account to use Google Colab and Google Drive. Basic programming knowledge using Python 3
If you have no programming background, you may still take the course. You can still benefit from the mathematical understanding of the subject. You can play with all the code that is provided.
All codes are provided for you to download.
No software needed to install on your computer. We will use Google Colab with step-by-step set up instructions included in the course.

Description
If you are interested in Machine Learning, Neural Networks, Deep Learning, Deep Neural Networks (DNN), and Convolution Neural Networks (CNN) with an in-depth and clear understanding, then this course is for you.

Topics are explained in detail. Concepts are developed progressively in a step by step manner. I sometimes spent more than 10 minutes discussing a single slide instead of rushing through it. This should help you to be in sync with the material presented and help you better understand it.

The hands-on examples are selected primarily to make you familiar with some aspects of TensorFlow 2 or other skills that may be very useful if you need to run a large and complex neural network job of your own in the future.

Hand-on examples are available for you to download.

Please watch the first two videos to have a better understanding of the course.

TOPICS COVERED

What is Machine Learning?

Linear Regression

Steps to Calculate the Parameters

Linear Regression-Gradient Descent using Mean Squared Error (MSE) Cost Function

Logistic Regression: Classification

Decision Boundary

Sigmoid Function

Non-Linear Decision Boundary

Logistic Regression: Gradient Descent

Gradient Descent using Mean Squared Error Cost Function

Problems with MSE Cost Function for Logistic Regression

In Search for an Alternative Cost-Function

Entropy and Cross-Entropy

Cross-Entropy: Cost Function for Logistic Regression

Gradient Descent with Cross Entropy Cost Function

Logistic Regression: Multiclass Classification

Introduction to Neural Network

Logical Operators

Modeling Logical Operators using Perceptron(s)

Logical Operators using Combination of Perceptron

Neural Network: More Complex Decision Making

Biological Neuron

What is Neuron? Why Is It Called the Neural Network?

What Is An Image?

My “Math” CAT. Anatomy of an Image

Neural Network: Multiclass Classification

Calculation of Weights of Multilayer Neural Network Using Backpropagation Technique

How to Update the Weights of Hidden Layers using Cross Entropy Cost Function

Hands On

Google Colab. Setup and Mounting Google Drive (Colab)

Deep Neural Network (DNN) Based Image Classification Using Google Colab. & TensorFlow (Colab)

Introduction to Convolution Neural Networks (CNN)

CNN Architecture

Feature Extraction, Filters, Pooling Layer

Hands On

CNN Based Image Classification Using Google Colab & TensorFlow (Colab)

Methods to Address Overfitting and Underfitting Problems

Regularization, Data Augmentation, Dropout, Early Stopping

Hands On

Diabetes prediction model development (Colab)

Fixing problems using Regularization, Dropout, and Early Stopping (Colab)

Hands On: Various Topics

Saving Weights and Loading the Saved Weights (Colab)

How To Split a Long Run Into Multiple Smaller Runs

Functional API and Transfer Learning (Colab)

How to Extract the Output From an Intermediate Layer of an Existing Model (Colab), and add additional layers to it to build a new model.

Who this course is for
Who is this course for? Almost for everyone. Machine Learning is not a topic for one single profession. Machine Learning (along with neural networks) is an immensely powerful tool that may help you to find solutions to some of the problems that one may not know how to solve otherwise. Try this course and see if it gives you better insight to address some of the problems you are working on.
People from a diverse range of professions may find this knowledge useful in their own profession.
Topics are explained in detail. Concepts are developed progressively in a step by step manner. I sometimes spent more than 10 minutes discussing a single slide instead of rushing through it. This should help you to be in sync with the material presented and help you better understand it.
The hands-on examples are selected primarily to make you familiar with some aspects of TensorFlow 2 or other skills that may be very useful if you need to run a large and complex neural network job of your own in the future.
Please watch the first two videos to have a better understanding of the course.