Udemy – AI & Deep Learning From Scratch In Python 2024-2

AI & Deep Learning From Scratch In Python Course. This course provides a comprehensive understanding of Convolutional Neural Networks (CNNs) by combining mathematical foundations and practical implementation in Python. Designed for beginners with no specific background, this course covers the fundamental concepts of Python programming and the computations required for deep learning. Each component of a convolutional neural network is first explained mathematically and then implemented step-by-step in Python. Through interactive programming exercises that can be run directly in the browser environment, participants gradually develop a complete object recognition framework based on an optimized CNN model. One of the class achievements of this course is an introduction to one of the most advanced and effective real-time multi-object recognition algorithms. Understanding the post-propagation process from both theoretical and practical aspects provides a solid foundation in the fundamental aspect of neural network training. By the end of the course, students will gain practical experience in building a complex CNN framework including advanced optimization and regularization techniques, which will enhance their ability to solve complex real-world object recognition problems with outstanding performance results and enhance their skills in the fields of computer vision and deep learning.

What you will learn

• Practical and mathematical understanding of neural networks:
• Understand how Deep Neural Networks work, both practically and mathematically.
• Understand the processes of forward and backpropagation, both mathematically and practically.
• Design and implementation of models:
• Design and implementation of a deep neural network for multi-class classification.
• Understanding and implementing the components of convolutional neural networks.
• Understand and implement advanced optimization, regularization, and initialization techniques.
• Model training and practical application:
• Training and validating a convolutional model on widely used datasets such as MNIST and CIFAR-10.
• Understanding and implementing transfer learning.
• Using a convolutional model to create a real-time multiple object recognition system.

This course is suitable for people who:

• Anyone who is interested in truly understanding convolutional neural networks and wants to build their own Object Detection Framework in Python.

Course details

• Publisher: Udemy
• Instructor: Victor Huerlimann
• Training level: Beginner to advanced
• Training duration: 5 hours and 16 minutes