This course covers an introduction to artificial intelligence (AI) focusing on computational approaches that enable automatic decision making. This includes algorithms for modeling relationships, search, routing, assignment, tracking, and planning. Techniques from optimization theory, operations research, and machine learning will be framed in terms of numeric and symbolic formulations, with an emphasis on data and knowledge representation in the presence of uncertainty. The course will provide the foundation for creating autonomous agentic systems, possibly involving robotics, operating at different levels of supervision with emergent behavior from multi-agent and human interaction. Emphasis will be given to the issues of ethics, safety, and alignment.
Topics:
a. Search and optimization, distances, spaces, constraints, and data structures
b. Making discrete decisions in discrete worlds with logic and decision trees
c. Continuous decisions and linear programs
d. Applications of information retrieval, planning, and matching
e. Approximation and curve fitting, learning in loss and cost functions
f. Probability and risk, Bayes rule, independence, uncertainty, entropy, anomaly detection, information
g. Sequences, Markov models, causality
h. System identification, modeling, filtering, and tracking
i. Neural networks, convolutional neural networks, and processing images
j. Text mining and natural language processing
k. Reinforcement learning
The course introduces students to the underlying concepts and applications of artificial intelligence in engineering, providing an introduction to complement to courses focused on machine learning. At the completion of the course students will be able to perform the following:
1. Identify tasks that involve artificial intelligence and select appropriate techniques for solving.
2. Use an understanding of optimization, data structes, and data processing functions along with objectives to desribe how AI can create automatic decisions, solve complicated problems, search spaces, or create models using learning algorithms.
3. Write computer programs or optimization problem formulations to find solutions or create models for tasks on different domains including images and text.
4. Understand the limitations and drawbacks of AI including ethical concerns of AI.