[AI 1] is a math heavy course offered at KTBYTE, and require students to have mastered self-guided learning. Students will learn tools to model and understand complex data sets, tools and algorithms that are commonly used for tackling "Big Data" problems. Covered topics include different techniques in supervised learning, unsupervised learning and reinforcement learning. This course is taught in Python using the pandas, numpy, and sk-Learn libraries. Students will have roughly 2 hours of homework assignments per week, plus a final project due at the end of the semester. [AI 1] vs Core classes: [AI 1] provides the theoretical and mathematical foundations to understand learning, and students do regular problem sets. The goal is to derive and understand the actual equations of various models. This includes techniques such as clustering, linear regression, and naive bayes. For many KTBYTE students, [AI 1] is also the first time they program using python. Unlike core classes, students are not taught python 'from the ground up', and are expected to pick up the language as it is used with examples in class.Research from KTBYTE students and alumni
Completion of [CORE 6a] or AP CS, or permission of instructor. Also requires Algebra II math experience.
These are examples of projects that students create as they grow their skills in [AI 1]
Working With Data: Finding Statistics
Importing data sets and finding statistics
Working with Data: Slicing and Indexing
Slicing and indexing data sets
Classification: Types of Problems and Models
What types of problems and models exist in machine learning? What do most models have in common?
Regression: Linear Regression
Linear Regression and Feature Importance
Regression: Common Regression Problems
Types of regression models and what each one is typically used for.
Regression: Gradient Descent
How does gradient descent work, and how can we use it to optimize our models?
Classification: Logistic Regression
Classification: Decision Trees
Decision trees and feature importance
Classification: More Decision Trees and Random Forest
More on decision trees and using ensemble methods to improve performance
Clustering models and unsupervised learning
Train test split AUC score, accuracy / precision / recall
Finding/starting a project
Finding and starting a project
Related Works + Experiment Design
Writing, and related works
Writing, Introduction and Abstract
Finishing the Research Project