Intuitive Deep Learning (بالعربي)
The place to build your 𝐟𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐤𝐧𝐨𝐰𝐥𝐞𝐝𝐠𝐞 in Deep Learning, the way it's meant to be learned.
Who Is This Course For ?
If you’re looking for a shallow, lightweight course, this isn’t the right fit. Our curriculum is designed for those who want to dive into the foundations of Deep Learning and apply sophisticated concepts to real-world challenges. Many of our participants are PhDs or experienced engineers, reflecting the rigorous and scientific nature of the material.
In this course, we don't take anything for granted and build knowledge from the ground up. It is perfect for professionals and researchers from other scientific domains eager to build a robust foundation in Deep Learning. If you’re a critical thinker seeking a solid, comprehensive learning experience that pushes beyond the basics, then this is the course for you.
Course Content
-
-
Chapter Introduction
FREE PREVIEW -
Table of Content
FREE PREVIEW -
Linear Regression
FREE PREVIEW -
The Line of Best Fit
FREE PREVIEW -
The Normal Equation
FREE PREVIEW -
Linear Classifiers
FREE PREVIEW -
Probabilistic Modelling
FREE PREVIEW -
Maximum Likelihood Principle
FREE PREVIEW -
Cross Entropy Derivation
FREE PREVIEW -
MSE Derivation
FREE PREVIEW -
Gradient Descent
-
Gradient Descent Derivation
-
Local vs Global Minimum
-
Pure Optimization vs Machine Learning
-
LR Schedules, Momentum & ADAM
-
Principle Component Analysis
-
-
-
Table of Content
FREE PREVIEW -
Backpropagation
-
Derivation by Hand
-
Chain Rule of Calculus
-
Computational Graphs
-
Fully Connected Networks
-
Data Separation through Non-Linearity
-
Universal Approximation Theorem
-
Parameters Norm Penalty
-
Dropout Derivation
-
Parameters Initialization
-
Practical Setup
-
Hands-On: Gender Classification From Scratch
-
Introduction
-
DataLoader
-
Forward Propagation
-
Backpropagation
-
ADAM & LR Schedule
-
Training & Testing
-
Deep Learning Questions
-
-
-
Table of Content
FREE PREVIEW -
Convolution
-
Image Processing
-
Convolution in ML
-
Convolution vs Dense Layers
-
Convolutional Neural Networks
-
Convolutional Networks
-
Residual Connections
-
Batch Normalization
-
Transfer & Multi-Task Learning
-
Fully Convolutional Networks
-
Receptive Field
-
Encoder Decoder Architecture
-
Tricks for Efficient Design
-
UNet Architecture
-
-
-
Table of Content
FREE PREVIEW -
Object Detection
-
Introduction
-
IoU & NMS
-
Accuracy Metric
-
Precision & Recall
-
PR Curve & mAP
-
RCNN, Fast RCNN & Faster RCNN
-
YOLO v1 to v4
-
Augmentation, Loss Functions & Debugging
-
Data Augmentation Use Cases
-
Loss Functions
-
L1, L2 & Huber Loss
-
Focal Loss
-
Contrastive Loss
-
Debugging
-
Learning Curves
-
Stability Issues
-
Grad CAM
-
Memory & Speed
-
Hands-On: Semantic Segmentation with TF
-
Computer Vision Questions
-
.png)
About this course
- €49,99
- Text lessons
- Arabic video content
- A community to ask questions
-
Lesson 1Linear RegressionLearn how to formulate the problem mathematically and solve the normal equation.
-
Lesson 2Linear ClassificationUnderstand probabilistic modelling, maximum likelihood principle and derive the main loss functions.
-
Lesson 3Gradient DescentDerive gradient descent formula and understand the difference between optimization vs learning.
Exercise: Linear models from scratch vs sklearn. -
Lesson 4BackpropagationDiscover how backpropagation made deep learning possible and which problems did it solve.
-
Lesson 5Fully Connected NetworksVisualize how activation functions separate the data, what is model capacity and how to initialize the weights.
-
Hands-on 1Gender ClassificationMaster Python: write a clean code for data loaders, neural networks and ADAM from scratch.
-
Lesson 6ConvolutionUnderstand basic image processing and how it can be implemented with convolution.
-
Lesson 7Convolution Neural NetworksReplace dense layers with convolutions and design deep networks with batch norm and residual connections.
Exercise: CNN for gender classification. -
Lesson 8Fully Convolution NetworksUnderstand the notion of receptive field to design powerful image-to-image networks.
-
Lesson 9Object DetectionExplore the full history of object detection algorithms: RCNN vs YOLO.
-
Lesson 10Augmentation, Loss functions & DebuggingExplore self-supervised learning using data augmentation and exotic loss functions and master debugging.
-
Hands-on 2Image SegmentationMaster TensorFlow: implement an industrial code, explore different segmentation tasks and design a powerful UNet
Still hesitating? Don't take my words for it
Highly Recommended Learning Journey
Safaa ALMOKDAD
I have not yet finished the course; however, I am happy with what I have learned so far. Issa explains the most challenging concepts in an easy-to-understand manner. You can follow along with a textbook to take notes. One notable aspect is that...
Read MoreI have not yet finished the course; however, I am happy with what I have learned so far. Issa explains the most challenging concepts in an easy-to-understand manner. You can follow along with a textbook to take notes. One notable aspect is that he explains Deep Learning, starting from the very first step in DL to the last one. Although I have been enrolled in a machine learning program before, I always lacked knowledge of DL as Issa presents in his course. While following the course, you will discover and understand some of the things that were ambiguous to you. The course includes quizzes that you can take to test your knowledge and hands-on exercises to apply what you have learned so far. Another interesting aspect is the application in this course; it focuses on working with images. By the end, you will gain experience in computer vision, presented by someone who has worked extensively in this field. The spoken language of the course is Arabic, so it's a plus point for the Arabic community."
Read LessA Highly Recommended Blend of Theory and Practical Expertise
Ali TFAILY
Issa's course, in my opinion, is exceptional, especially for Arabic speakers. It represents the essence of years of concentrated expertise, condensing key concepts that Issa skillfully presents to streamline the learning path for students, offeri...
Read MoreIssa's course, in my opinion, is exceptional, especially for Arabic speakers. It represents the essence of years of concentrated expertise, condensing key concepts that Issa skillfully presents to streamline the learning path for students, offering what can be challenging to find in various courses today. Notably, alongside well-explained scientific concepts, this course incorporates valuable insights derived from Issa's experience, including work on previous projects. This distinctive feature makes the course stand out. I believe that with its blend of ideas and practical applications, this course will provide a much deeper understanding to those who study it, surpassing their prior knowledge. I highly recommend it!
Read LessA Great Start in the World of Deep Learning
Hussien Ismail
I came across this course as an attempt to enter into the world of deep learning and integrate it into the field ofstructural design and topology optimization. I started with zero knowledge about the topic, and now I have the solid foundation need...
Read MoreI came across this course as an attempt to enter into the world of deep learning and integrate it into the field ofstructural design and topology optimization. I started with zero knowledge about the topic, and now I have the solid foundation needed to apply what I've learned to my field. The best part of the course is that Eng. Issa explains the content as if "I am 5 years old," making it the perfect way to absorb and remember each topic. I highly recommend this course, especially for anyone willing to pursue multidisciplinary research that integrates deep learning.
Read LessCheck our content curated from the course
Who is behind the course?
Issa Hammoud
Senior Machine Learning Scientist
I am a Senior Machine Learning Scientist and University Instructor with several years of experience. I have worked in France with startups and big companies, in R&D and production, to solve real world problems.
I like to understand the intuition behind science, and I have a good taste of coding.
Frequently Asked Questions
Reach out to us at [email protected] for any specific inquiries.
-
What sets this course apart from other online courses?
Unlike many other courses that offer lightweight content, this course provides a comprehensive summary of key books and papers. We delve into the reasoning behind techniques, present mathematical derivations, and use engaging animations—all designed to help you grasp the underlying intuition.
-
Is this course only for Arabic speakers?
While the video content is currently only in Arabic, everything else, including text lessons, hands-on activities, interview simulations, and community access, is in English. So, you can still benefit from the course even if you are an English speaker.
-
What resources were used to create this course?
This course was primarily inspired by Ian Goodfellow's book "Deep Learning" for foundational concepts, various original research papers, and my own professional experience.
-
What is the estimated time to complete this course?
Depending on your level and consistency, the course should take between 1 and 2 months to complete. We've designed it to be compact to maximize your efficiency, allowing you to gain knowledge that will benefit your professional life in a relatively short time frame.