Instructor Information

Nathaniel D. Bastian

Nathaniel D. Bastian, PhD is a leader, practitioner, researcher, and educator of mathematical, computational, analytical, data-driven, and decision-centric methods to support the improvement and enhancement of decision-making in cyber security, national defense, military operations, human resources and manpower, healthcare, logistics, energy and finance. He is a decision analytics professional with expertise in the scientific discovery and translation of actionable insights into effective decisions using algorithms, techniques, tools and technologies from operations research, data science, artificial intelligence, systems engineering, and economics to research, design, develop, and deploy intelligent decision-support models, tools and systems for descriptive, predictive and prescriptive analytics. He has authored over 60 refereed journal and conference papers, several book chapters, and one textbook. He is the recipient of numerous academic awards, honors and grants, to include a Fulbright Scholarship and National Science Foundation Graduate Research Fellowship. He serves as an Associate Editor for four journals, as well as Referee for over 20 journals. He is an active member of MORS, INFORMS, ACM, IEEE, SIAM, and AAAI.

DISCIPLINARY EXPERTISE

- Optimization, simulation, statistical computing, machine/deep learning, intelligent systems, big data analytics

- Decision science, business analytics, applied econometrics, production economics, engineering management

RESEARCH INTERESTS

- Computational stochastic optimization and robust learning for making inferences and decisions under uncertainty

- Multiple objective optimization and federated machine learning for distributed resource allocation decision-making

Course Information

Course Description

In order to drive a future where artificial intelligence (AI) enabled autonomous systems are trustworthy contributors to society, these capabilities must be designed and verified for safe and reliable operation and they must be secure and resilient to adversarial attacks. Further, these AI enabled autonomous systems must be predictable, explainable and fair while seamlessly integrated into complex ecosystems alongside humans and technology where the dynamics of human-machine teaming are considered in the design of the intelligent system to enable assured decision-making. In this course, students are first introduced to the field of AI, covering fundamental concepts, theory, and solution techniques for intelligent agents to perceive, reason, plan, learn, infer, decide and act over time within an environment often under conditions of uncertainty. Subsequently, students will be introduced to the assurance of AI enabled autonomous systems, including the areas of AI and autonomy security, resilience, robustness, fairness, bias, explainability, safety, reliability and ethics. This course concludes by introducing the concept of human-machine teaming. Students develop a contextual understanding of the fundamental concepts, theory, problem domains, applications, methods, tools, and modeling approaches for assuring AI enabled autonomous systems. Students will implement the latest state-of-the-art algorithms, as well as discuss emerging research findings in AI assurance.

Course Goal

To introduce the field of AI along with the fundamental considerations for assuring AI enabled autonomous systems and then apply that knowledge to prepare groups to implement the applicable algorithms, methods, tools and modeling approaches to design and build an assured AI model/system by the end of the semester.

Course Objectives

  • Explain the fundamental concepts, theory, problem domains, applications and solution techniques spanning the field of artificial intelligence.
  • Apply the latest algorithms, methods, tools and modeling approaches for designing and building AI enabled autonomous systems.
  • Identify emerging, state-of-the-art algorithms useful for building secure, resilient, robust, fair, unbiased, explainable, safe, and reliable AI systems.
  • Describe the dynamics of human-machine teaming in the context of building trustworthy, predictable and assured AI enabled autonomous systems, as well as communicate the policy considerations and ethical principles associated with developing and deploying responsible AI and autonomy.

When This Course is Typically Offered

This course is typically offered in online format in the fall and spring terms at JHU.

Syllabus

  • Introduction to Artificial Intelligence and Agent Architectures
  • Intelligent Search and Optimization for Reasoning
  • Supervised Machine Learning
  • Reasoning with Uncertainty
  • Planning with Uncertainty
  • Learning with Uncertainty
  • Multiagent Systems
  • Learning to Act
  • Assurance of AI Enabled Autonomous Systems
  • AI System Security, Resilience and Robustness
  • AI System Fairness and Bias
  • AI System Explainability
  • Human-Machine Teaming
  • AI System Safety, Reliability and Ethics

Student Assessment Criteria

Discussion Assignments 6%
Programming Assignments 60%
Group Project 34%

Assignments are due according to the dates posted in the Blackboard course site. You may check these due dates in the Course Calendar or the Assignments in the corresponding modules. I will post grades in the Blackboard grade book as soon as I have finished grading all students. I generally do not directly grade spelling and grammar. However, egregious violations of the rules of the English language will be noted without comment. Consistently poor performance in either spelling or grammar is taken as an indication of poor written communication ability that may detract from your grade. A grade of A indicates achievement of consistent excellence and distinction throughout the course—that is, conspicuous excellence in all aspects of assignments and discussion in every week. A grade of B indicates work that meets all course requirements on a level appropriate for graduate academic work. These criteria apply to both undergraduates and graduate students taking the course.

Computer and Technical Requirements

You will need access to Python 3.8 (or newer). I recommend that you use the Anaconda Individual Edition, which is the most widely used Python distribution platform, as you will be expected to use Jupyter notebooks as part of your programming assignments.

Participation Expectations

This course is conducted entirely online via Blackboard. This course is divided into 14 modules that are completed over the same number of weeks; each week starts on Monday and ends on Sunday. There are no required class meeting times for the course; however, optional synchronous Office Hours will be offered. In addition, smaller sized Critical Thinking Groups may need to “meet” periodically to collaborate using a dedicated online discussion group within Blackboard, Zoom, or other methods. Each week, you are required to complete the assigned readings, view the lecture recordings with associated notes, watch the supplemental videos, and participate in the discussion board. You are encouraged to attend the optional synchronous office hours each week. You demonstrate mastery of the course concepts by completing several programming assignments, as well as conducting a final group project with written report and recorded presentation. You are expected to complete all deliverables by their assigned due dates.

Textbooks

Textbook information for this course is available online through the MBS Direct Virtual Bookstore.

Course Notes

There are notes for this course.

Final Words from the Instructor

Upon completion, my hope is that you become a better engineer, scientist, analyst, etc. with deeper knowledge of assured artificial intelligence and autonomy.

Term Specific Course Website

http://blackboard.jhu.edu

(Last Modified: 06/25/2021 06:36:53 AM)