Code: AI41F

Type

Fundamental

For Grades

9-12

Target Competitions

This course aims to inspire in machine learning, provide project ideas, prepare students for the STEM project competitions both in theory and in practice. The level of the projects in the course is prestigious national / international competitions such as:

• Regeneron ISEF
• The Siemens Competition in STEM
• Broadcom MASTERS competition

Description

This course introduces deep learning using PyTorch. Students will learn how deep neural networks work, train models for computer vision and natural language processing tasks, and build real AI applications. The course covers CNNs for images, Transformers for text, and emphasizes understanding fundamentals while building practical projects. By the end of this course, students will be able to:

• Understand how deep neural networks learn from data
  
• Build and train neural networks from scratch using PyTorch
  
• Implement the complete PyTorch training loop (forward, loss, backward, optimize)
• Design and train Convolutional Neural Networks (CNNs) for image tasks
• Work with pre-trained models and transfer learning
• Understand and apply Transformer models for text tasks
• Process and classify text data using BERT and GPT-style models
• Understand advanced concepts: data augmentation, regularization, optimization
• Build complete AI applications: image classifiers, object detectors, text classifiers, chatbots
• Debug and improve deep learning models
• Choose between vision and NLP approaches for different problems

Who should take this course?

This course requires approval after registration. The minimum requirements for this course is as follows:

• AI: Introduction to Machine Learning
• Math: Algebra II

Competition background is not required but is a plus: USACO Gold, AIME or above

Content

The course is composed of 4 modules:

• Neural Network Fundamentals
• Building Better Networks
• Computer Vision with CNNs
• NLP & Transformers + Final Project

---