Deep Learning 101 with Python and Keras

Deep Learning Doesn’t Have to Be Confusing

Have you heard of these technologies—but have no idea how they work?

Deep learning can sound incredibly complicated. For example to understand CNNs you need to at least understand this sentence:

"This architecture has some building blocks that you already know, such as fully connected layers and sigmoid activation functions, but it also introduces two new building blocks: convolutional layers and pooling layers." From Hands-on Machine Learning and Scikit-Learn, Keras & Tensorflow

If you have a good understanding of basics of neural networks, though, this sentence would be a piece of cake to decipher.

Deep Learning 101 will teach you all the essentials of deep learning without wasting your time.

Here is how we're going to do that

Explanation of key concepts

The main feature of the course is the video explanation of key concepts of deep learing. We will start from simple topics such as what neural networks are and how they learn and build our knowledge up to RNNs and CNNs.

Implementation of each technique

All of the concepts we introduce in this course are things we can implement using Keras and Python. After each theoretical lesson, there will be videos showing you how to implement each of the concepts we introduced in the theoretical section.

Guided exercises

At important milestones of your learning, there will be video walk throughs of exercises. These include importing and preparing datasets, building and compiling neural networks, applying regularization to networks, evaluating and comparing networks and more.

Programming assignments

After the exercises, you will be given assigments to interact with the code from the exercises in order to understand some of the key concepts better.

Lesson slides

You will have access to all the material used while teaching. The course slides will be shared with each lesson for you to keep a copy.

End-of-chapter quizzes

At the end of each chapter, you will have a couple of questions to answer. The answers of these questions will also be shared with you. Even though all of the topics discussed during the lessons are important, these questions will emphasize the must-know points from each lesson.

Lesson summary notes

Each lesson comes with its own PDF of summary notes. These notes will include key information that was introduced during the classes. You can download and keep these notes to use as a reference during the exercises or in the future.

Final project

At the end of the course, we will do a final project using a pre-trained model from the internet to solve an image recognition task. This project will help you understand how deep learning is used in the industry.

Access to instructor through comments

If you have any questions or comments, you can leave a comment on lesson videos or exercises and your instructor will be

Course repository

You will have access to all the code developed in the course. You can use this repository to follow exercises along, do your course assignments but also in the future as a reference to remember how to implement any of the concepts you learned during the course.

Ella Farber
“I have tried other e-learning courses on this subject but none could really explain it to me. Mısra’s course finally made me understand data science.”

This course is for you if:

  • you know beginner level data science
  • you have not had a traditional education on AI
  • you are willing to put in the work
  • you want to be able to follow the latest developments in AI like GPT-3
  • you want to improve the work you do by applying or experimenting with deep learning
  • you are interested in practical uses of deep learning

This course is not for you if:

  • you want to take one course and become a deep learning expert
  • you don't want to put in the work to get your essentials straight
  • you want to learn all the math and theory behind deep learning algorithms

Coming soon!

Deep Learning 101 will launch on 18 December 2021.

Leave your email below to be notified.

Get updates!

On this page, I will post updates on the progress of the course. I will also share these updates through email to those interested. If you'd like to get these updates and more useful *free* content from me, sign up below:

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Enroll now!

Regular edition

  • 13 modules explaining DL concepts
  • Assignment, quizzes and hands-on exercises
  • Lesson summary notes
  • Access to course code repository
  • Access to your instructor through comments
  • Hyperparameters cheat sheet
  • One 60-minute video call to answer your questions one-on-one
Buy now for $99
Sales tax or VAT might apply.

Premium edition

  • 13 modules explaining DL concepts
  • Assignment, quizzes and hands-on exercises
  • Lesson summary notes
  • Access to course code repository
  • Access to your instructor through comments
  • Hyperparameters cheat sheet
  • One 60-minute video call to answer your questions one-on-one
Buy now for $199
Sales tax or VAT might apply.

Get 20% off during launch week!

30-day money-back guarantee

This course comes with a 30-day money-back guarantee. If you’re not happy with the course for any reason, I'll refund your payment in full.

Here's what we will cover in this course...

Introduction to Deep Learning

Building blocks of Deep Learning

Exercise - Let's build a simple Neural Network

Hyperparameters of Neural Networks

Over/underfitting problem and solutions

Unstable Gradients

Computational time

Exercise - Setting up neural networks

Model diagnosis and making improvements

Hyperparameter tuning

Exercise - Improving the model

Final Project - Building a web app

Module 1

Introduction to Deep Learning

  • Quick overview of deep learning
  • What are neural networks
  • Common DL techniques

Module 2

Building blocks of Deep Learning

  • Weights and biases
  • Activation function
  • Vectorization of calculations
  • Forward propagation
  • Backward propagation
  • Gradient descent

Module 3

Exercise - Let's build a simple Neural Network

  • Setting up a virtual environment
  • Finding, importing and preparing data
  • Building a network
  • Model evaluation

Module 4

Hyperparameters of NNs

  • Overview of NN working principles
  • Number of neurons
  • Number of hidden layers
  • Loss functions and optimizers
  • Activation functions
  • Batch size
  • Regularization
  • and more...

Module 5

Over/underfitting problem

  • What is overfitting?
  • L1 and L2 regularization
  • Dropout regularization
  • Data augmentation
  • Early-stopping

Module 6

Unstable gradients problem

  • What is vanishing/exploding gradients?
  • Weight initialization techniques
  • Non-saturating activation functions
  • Batch normalization
  • Gradient clipping

Module 7

Computational time

  • Data normalization
  • Batching
  • Optimization algorithms
  • Network pruning
  • Learning rate scheduling

Module 8

Exercise - Setting up NNs

  • Initializing the model
  • Dealing with underfitting
  • Dealing with overfitting
  • Experimenting with changing hyperparameters

Module 9

Model diagnosis and making improvements

  • Model evaluation metrics
  • Levels of performance
  • Determining the performance level to aspire to
  • Making a plan for improvements

Module 10

Hyperparameter tuning

  • Why is this a problem?
  • Common ways to tune hyperparameters
  • Advanced hyperparameter tuning algorithms

Module 11

Exercise - Improving the model

  • Detecting the performance gap
  • Setting up improvements

Module 12

Final Project - Building a web app

  • Where to find pre-made models
  • Integrating pre-made models
  • Making a web app for our project

Hi, I am Mısra...

I will be your instructor on this course.

Over the years, as a data scientist, I worked at big companies like IBM and also small niche companies that are trying to make the world a better place. Now, I teach what I learned about data science online.

As a computer science student, I never got the formal education on deep learning. When I was studying, deep learning was not even a common topic that was taught at school. Still, I wanted to use deep learning to have a robot choose its own hand and arm gestures to accompany its speech using RNNs. And then my journey into the deep realm of deep learning began.

It took me years of self-study, trial and error, many books and online courses to come to the point where I am right now. My goal is to bring you to that same point without all the effort and time it took me.

I had to learn deep learning from many resources that teach things in the most formal way possible. Burying its students with endless equations, math formulas and theoretical, abstract concepts before being able to understand the bigger picture logic. If you take a practical approach though it becomes much easier to learn deep learning. That's what we're going to do in this course.

This is an online course you will not abandon after a couple of days. The material is advanced but the delivery will help you feel progress every step of the way.

Looking forward to seeing you in there!

Ella Farber
“I have tried other e-learning courses on this subject but none could really explain it to me. Mısra’s course finally made me understand data science.”

Frequently Asked Questions

Q: What are the requirements for taking this course?

You need to have an understanding of basic data science concepts such as training a model, testing a model, evaluation metrics, classification vs. regression, supervised vs. unsupervised problems.

Everything else, we will cover in this course.

Q: Why should I prefer this course over other online courses?

The difference of Deep Learning 101 from other online courses is the delivery and the teaching approach. In this course, the goal is to bring you from very little knowledge on data science to having a solid grasp on all essential deep learning topics. At the end of this course, you will have a full understanding of all the building blocks and this will enable you to understand more complex topics and algorithms that were built using these building blocks.

We will not overload you with information that will not stay with you after a couple of days. The goal is not to throw every little detail at you but rather to give you distilled information that will enable you to apply deep learning in whichever way you need it.

The lessons are not in a university class format like in many other online courses. For every new concept we start from a high-level working logic and only if necessary go into the depths of theory.

Q: Why should I take this course when there are so many free online resources?

Of course, there are many free resources online about data science and deep learning. The main advantage of taking Deep Learning 101 is to have someone to collect, distill and organize the information for you in a way that is ready to be consumed right away.

Learning by yourself costs time and extra effort to find just the right information that is not outdated. And it can become an overwhelming and tiresome exercise.

I have been through that already and now, I offer you a shortcut.

Q: Which programming language is used in exercises?

The exercises are done in Python in a Jupyter notebook using Keras with a Tensorflow backend. All of these technologies will be introduced during the course but there is not Python or Jupyter notebook tutorial inside the course.

Q: Do I need excellent math skills to join this course?

You do not need to have any prior math knowledge to start this course.

Q: How long will this course take to finish?

If you take the time to watch all the lessons and do all the exercises, the course will take you around a month to complete, assuming you have other engagements that you work on full-time.

Q: Which OS (operating system) is the course suitable for?

The course is OS agnostic. We will be setting up and and executing the exercises in MacOS but as long as you have access Jupyter Notebooks, that should not be a problem. We will also introduce a way to follow the exercises in case you have a slow computer that cannot do the computations locally.

Q: Can I get a refund if the course does not meet my expectations?

Yes. This course comes with a 30-day money back guarantee. If, for any reason, the course does not meet your expectations and you would like a refund, send me an email at misra@soyouwanttobeadatascientist.com and I will arrange your refund.

Enroll now!

Regular edition

  • 13 modules explaining DL concepts
  • Assignment, quizzes and hands-on exercises
  • Lesson summary notes
  • Access to course code repository
  • Access to your instructor through comments
  • Hyperparameters cheat sheet
  • One 60-minute video call to answer your questions one-on-one
Buy now for $99
Sales tax or VAT might apply.

Premium edition

  • 13 modules explaining DL concepts
  • Assignment, quizzes and hands-on exercises
  • Lesson summary notes
  • Access to course code repository
  • Access to your instructor through comments
  • Hyperparameters cheat sheet
  • One 60-minute video call to answer your questions one-on-one
Buy now for $199
Sales tax or VAT might apply.