AI Competitions are a key way for students to showcase their skills, build accomplishments for college and develop a lifelong love of learning and practicing AI. This course introduces students to the structure of AI competitions and how to compete in such competitions.
During the course, all students will focus on their entry to a Kaggle competition - one of the industry’s most prestigious AI competitions. The camp will cover techniques to optimize data, algorithms and teach students how to assess their entries. Students can use any combination of the techniques to improve their scores individually, and can submit their entries after the camp. They will also have a lifetime access to their AIClub accounts and can continue to work on their entries after the camp if they wish.
Curriculum
We will review different types of AI for numbers, text, categorical data and images and the basics of algorithms for each. We will then introduce new concepts of data tuning (called Feature Engineering, Feature Optimization) and show students how to assess their datasets and improve their AIs through a combination of data optimization and algorithm optimization.
They will build several AIs reflecting improving scores for their chosen project. At the end of the camp, they will have a competition entry that they can choose to submit or work on further. This knowledge will equip them to use AI in other projects, such as STEM competitions. They will also get a good understanding of the relationship between AI quality and data quality and can use this knowledge for future real world projects.
Students will showcase their project to their parents and peers on the last day of the summer camp.
The core AI concepts they will learn are:
• Review different types of algorithms (classification, regression) appropriate for different types of problems and data (text, numbers, categories and images).
• Review how AI algorithms work (Linear Regression, Decision Trees, and Neural Networks) and how to tune and improve the AI performance.
• How to improve an AI with better data. Review core concepts of Data Skew, Balance, Overfitting etc. Learn new concepts of Feature Engineering, Normalization, Feature Optimization.
• How prestigious AI competitions work and how to engage in these competitions.
• How to train competition grade AIs.
Prerequisites
• A laptop with chrome browser
• Wifi connectivity
• Coding knowledge optional. No previous coding experience is okay.
• No previous math or algebra knowledge needed
• Having taken AI Basics (M1 or H1) or one of our introductory summer camps.
What Students Take Away
• A good understanding of AI - a fascinating and fundamental technology that is changing our world.
• The core skills needed to apply their AI knowledge in a competitive environment and showcase their skills. A path of lifetime learning and self improvement.
• A lifetime access to an online cloud account where they can continue to build new projects and learn more AI
• Opportunities to compete and win in AI competitions