Core Concepts in AI

Overview

The course "Core Concepts in AI" provides a comprehensive foundation in artificial intelligence (AI) and machine learning (ML), equipping learners with the essential tools to understand, evaluate, and implement AI systems effectively. From decoding key terminology and frameworks like R.O.A.D. (Requirements, Operationalize Data, Analytic Method, Deployment) to exploring algorithm tradeoffs and data quality, this course offers practical insights that bridge technical concepts with strategic decision-making. What sets this course apart is its focus on balancing technical depth with accessibility, making it ideal for leaders, managers, and professionals tasked with driving AI initiatives. Learners will delve into performance metrics, inter-annotator agreement, and tradeoffs in resources, gaining a nuanced understanding of AI's strengths and limitations. Whether you're a newcomer or looking to deepen your understanding, this course empowers you to make informed AI decisions, optimize systems, and address challenges in data quality and algorithm selection. By the end, you'll have the confidence to navigate AI projects and align them with organizational goals, positioning yourself as a strategic leader in AI-driven innovation.

Syllabus

Course Introduction

This course provides a comprehensive introduction to key concepts in artificial intelligence (AI) and machine learning (ML). Learners will explore essential vocabulary, the R.O.A.D. Framework, performance evaluation, and algorithm tradeoffs. Topics include data quality, inter-annotator agreement, and the strengths and weaknesses of AI methods. By the end, learners will be equipped with the foundational knowledge to navigate and assess AI and ML systems effectively.

Introduction to Artificial Intelligence

This module provides an introduction to artificial intelligence (AI). It does not require any prior knowledge of AI and is suitable for briefing managerial, and non-technical leaders to improve knowledge, expectations, and communication for AI projects.

Machine Learning

This module covers the statistical foundations of machine learning and the common metrics for evaluating machine learning and artificial intelligence performance.

Algorithm Tradeoffs

This module introduces the most common algorithms used in AI and machine learning, including support vector machines, Naïve Bayes, decision trees, random forest, and neural networks. We will discuss the strengths and weaknesses of these algorithms for different classes of problems.

Data

This module explores data types (nominal, ordinal, categorical) and the challenges of data labeling, including human cognitive limits and reference issues. A key focus is inter-annotator agreement—a method to measure labeling consistency, highlighting biases and inefficiencies in human and machine processes. Consistent labeling, often more impactful than advanced algorithms, is crucial for responsible AI.

Resources

This module introduces the most common resource considerations in AI, specifically memory, computational tradeoffs, query expressiveness, and algorithm performance.