• Overview
• Dual Enrollment with Bachelor’s Degree
  • 4+1 Model Tracks for AE, BME, CE, CS, CVE, EE, SE
  • General 4+1 / Early Entry Policies Including Tuition
  • 8 MB PowerPoint Overview Presentation on MSOE ML Programs
• Official Websites
  • Catalog
  • Academics
  • Admissions
  • Landing / Information Request
• Accreditation
• Machine Learning Careers In The News
• Direct Admit Requirements (also apply to Graduate Certificate in Applied Machine Learning)
• Curriculum
  • Required Courses
  • Example Plan
  • Advising
  • Grade Replacement
• Notes on Specific Courses
  • Quarter System Applied ML Graduate Certificate (2021–’22 and 2022–’23)
  • Graduate Certificate in Advanced Business Strategy Using AI and Analytics
  • CSC5201 Microservices & Cloud Computing
    • Substitutions Allowed: Graduate Courses
    • Substitutions Allowed: Undergraduate Courses
      • Acceptable Web Apps Coursework
      • Acceptable Cloud Computing Coursework
  • CSC5601 Theory of Machine Learning: CS Major and MSML Elective
  • CSC5980 Special Topics: An Engineering Approach to Game Theory and Economics
  • CSC6605 Machine Learning Production Systems: Alternate Prerequisites
  • CSC67XX
  • CSC6712 Distributed Storage Systems
  • CSC7901 Machine Learning Capstone
    • Enrollment Steps
    • Capstone Project Options
    • Potential Capstone Project Advisors
    • Capstone Project Process
    • Final Report
  • PHL6001 AI Ethics and Governance
• Course Details
• Past and Planned Offerings

Overview

The MS in Machine Learning program was first offered in fall of 2023, coinciding with MSOE’s transition to a semester system. It consists of 8, 4-credit courses.

All required courses are available online (fully virtual or in classroom with technology enabling remote participation). Many courses will be synchronous. The 4 classes that are part of graduate certificates (see track below) will be offered in the evening.

Students who were in the 2021–22 cohort of our applied ML graduate certificate, upon which the core classes in the MSML are based, were generally working full-time and said that it was better for them to take 1 course at a time instead of 2. So, for the master’s degree, that would translate to 3+ years depending on how many summer classes are taken. The program could be done in 2 years or a bit less if someone took 2 courses at a time.

We expect that many of the students in the program will have tuition support from their employers. Depending on the company, taking one course at a time might also be the best path from a financial perspective if the employer has annual limits on tuition support.

Note that the MSML program (and our ML graduate certificates) do not require letters of recommendation, although they are listed as optional on MSOE’s common application. Admission is strictly based on academic background and we personally reach out to applicants as needed to clarify any academic background that may be missing.

Please contact MSML program director Dr. Durant with any questions.

Dual Enrollment with Bachelor’s Degree

The MS can be taken as a 5-year early entry dual enrollment program with many MSOE undergraduate degrees including all CSSE (CS and SE) and ECBE (BME, CE, and EE) undergraduate degrees, and AE and CVE. Students must apply while they have junior standing, but should let their advisor know of their intent as early as possible.

• One or more summer courses are usually required.
• BMEs entering in F23 or later should choose the “AI Focus” option.
• All CSSE and ECBE undergraduate majors allow their free and program elective requirements to be met by MSML courses. Check with your advisor if not in one of these departments.
  • MSOE allows a maximum of 16 credits of graduate courses (i.e., 4 MSML courses) to be taken before the BS degree is conferred.
  • MSML courses meeting an undergraduate requirement count toward full-time undergraduate enrollment.
  • Normal undergraduate tuition applies prior to completing the undergraduate degree requirements.
• Juniors: apply here

4+1 Model Tracks for AE, BME, CE, CS, CVE, EE, SE

General 4+1 / Early Entry Policies Including Tuition

8 MB PowerPoint Overview Presentation on MSOE ML Programs

Official Websites

Catalog

• https://catalog.msoe.edu/. Select the latest Graduate Catalog.

Academics

• MS in Machine Learning
• Graduate Certificate in Machine Learning
• Graduate Certificate in Machine Learning Engineering - offered in fall, 2023; site coming soon

Admissions

• Machine Learning Programs
• MS in Machine Learning
• Graduate Certificate in Machine Learning
• Graduate Certificate in Machine Learning Engineering - offered in fall, 2023; site coming soon

Landing / Information Request

• MS in Machine Learning
• Graduate Certificate in Machine Learning
• Graduate Certificate in Machine Learning Engineering - offered in fall, 2023; site coming soon

Accreditation

Please see MSOE’s Accreditations for information accreditation at the university and program levels, including accreditation of all undergraduate programs in the CSSE Department. Note that as of 2023 ABET does not offer offer accreditation for masters programs in any computing disciplines.

Machine Learning Careers In The News

• 2023-12-22 Milwaukee Business Journal / MSOE ML Programs Win Wisconsin Inno Fire Award (PDF)
• 2023-08-14 Wall Street Journal article on demand for AI experts
• 2023-07-28 MSOE post on ML Careers and Industry Growth
• 2022-08-19 MSOE Post for Employers on Developing ML Talent

Direct Admit Requirements (also apply to Graduate Certificate in Applied Machine Learning)

• Technical bachelor’s degree
• Programming experience with a modern language such as Python, C#, or Java. One year of college coursework (culminating in e.g., CS2852, CSC1120, or CSC2611 or equivalent minimum.
  • Students with less experience (similar to CS1021, CSC1110, or CSC1310) will be encouraged to take CSC5120 Software Development for Machine Learning over the summer. This can count as an elective in the Master of Science in Machine Learning.
• At least a year of differential and integral calculus required. Multivariable calculus or linear algebra preferred.
  • Students who don’t have both topics will take MTH5810 Mathematical Methods for Machine Learning; students who do will take an additional elective.

Curriculum

Required Courses

The 8 required courses comprise 2, 2-course subsets that each meet the requirements for a graduate certificate, plus 4 additional courses. Key prerequisites are noted here; more detailed prerequisites are noted below and in the graduate catalog.

• Subset 1: Applied Machine Learning Graduate Certificate
  • CSC5610 AI Tools and Paradigms
  • CSC6621 Applied ML (prerequisite of CSC5610)
• Subset 2: Machine Learning Engineering Graduate Certificate
  • CSC5201 Microservices & Cloud Computing
  • CSC6605 ML Production Systems (prerequisite of CSC5610)
• MTH5810 Math Methods for ML
• PHL6001 AI Ethics and Governance
• CSC5xxx Elective (can be met with CSC5120 if this course is needed)
• CSC7901 Capstone

Early entry students should take both MTH2130 Calculus III and MTH2340 Linear Algebra with Applications and replace MTH5810 with an approved elective.

MSOE CS students and any student who has completed the data science sequence through CSC2621 replace CSC5610 with an approved elective.

MSML students may have either or both of the certificates conferred when they meet the course requirements by selecting “Completion of Certificate Program” on the registrar’s website. MSOE early entry students are generally not eligible for this since MSOE does not allow a course to be counted for more than 2 credentials and these students are typically using the MSML courses to also meet an undergraduate degree requirement.

Example Plan

Advising

As each student enters the program, the program director prepares a draft advising plan of which semester the student will take each course. Please contact the program director if you need a copy of your plan, and to make changes, such as increasing or decreasing how many classes you plan to take each semester.

Grade Replacement

MSOE’s machine learning programs automatically grant requests to retake and grade replace classes. This means that students in machine learning programs may re-enroll in a class without explicit permission. It also means that, if a grade below a B was earned the first time, the second grade can replace the original grade in GPA calculations. To take advantage of this option, students must notify their program director or coordinator. They will forward the request to the registrar’s office, which will then process the replacement at the end of the term. This must be done between week 2 and the week before exam week, preferably early in the term.

Notes on Specific Courses

Quarter System Applied ML Graduate Certificate (2021–’22 and 2022–’23)

Completing this certificate at MSOE (CS6230, CS6330, CS6340) meets the CSC5610 and CSC6621 requirements by substitution, reducing the number of additional courses needed to earn the MSML.

Graduate Certificate in Advanced Business Strategy Using AI and Analytics

The combination of BUS6121 Data Wrangling and Exploration and BUS6131 Predictive Analytics meets the CSC5610 requirement by substitution in the MSML.

The 3 credits in BUS6141 Analytics Leadership and Strategy can be used as MSML elective credits.

CSC5201 Microservices & Cloud Computing

Substitutions Allowed: Graduate Courses

There are opportunities to substitute certain courses for the CSC5201 requirement in the MSML and in the ML Engineering Graduate Certificate. CSC5201, CSC6711, or CSC6712 can satisfy the CSC5201 requirement. If MSML students take more than 1 of these classes, the additional courses are counted as MSML electives.

Substitutions Allowed: Undergraduate Courses

Students with sufficient coursework in both web apps and cloud computing replace this course with an approved elective.

Acceptable Web Apps Coursework

• SE2840 Web Application Development, Required for CS and SE
• SWE2511 Web Apps, Required for SWE

Acceptable Cloud Computing Coursework

• CS4230 / CSC4981 Distributed and Cloud Computing, Elective

Note: SWE2410 Design and Cloud Patterns is not sufficient to meet this requirement.

CSC5601 Theory of Machine Learning: CS Major and MSML Elective

Early entry CS majors have the option of taking this course to simultaneously satisfy their undergraduate requirement of CS3400 or CSC4601. Students who have credit for CS3400 or CSC4601 may not take CSC5601.

CSC5980 Special Topics: An Engineering Approach to Game Theory and Economics

For decades, the study of strategic interactions between rational decision makers has formed the basis of game theory. This course, using a two-volume textbook, extends game theory concepts to focus on dynamic games, introducing students to a new take on game theory referred to as the field theory of games. The course prioritizes conceptual understanding over mathematical equation solving, making the text accessible to not only engineering students but also to a more general audience, including business students.

By using a toolkit based on the Wolfram Language, students can bypass the need to solve linear programming problems and partial differential equations by hand, allowing them to arrive at solutions with practical applications more efficiently. An introduction to the Wolfram Language symbolic programming, Mathematica, is provided. Mathematica is used to study both empirical trends as well as engineering model simulations based on the engineering extension of game theory. A review of the concepts from physics and mathematics that underlie the approach is given. Along with these introductory foundations, economic applications will be presented and shown to depend on aspects of modern theories of differential geometry.

Though this course begins with classical game theory, it differs from the usual approaches to dynamic games and deals with incomplete information by using constraints in a geometric theory, where the shortest path provides a deterministic prediction of future behaviors. Students will learn to apply introductory ideas to a system without constraints. Next, students will explore the consequences of adding constraints and will be provided an application guide.

CSC6605 Machine Learning Production Systems: Alternate Prerequisites

Although CSC5201 is not a prerequisite for this class, students will benefit from having CSC5201 first whenever possible.

CSC2621 Introduction to Data Science (a course in the DS minor) is a sufficient prerequisite to take this course.

Also, CS and SE students will generally meet the following alternate prerequisites by the time they have early entry status.

CSC67XX

CSC67XX-prefixed courses are conducted in a combined seminar/project format. Instead of traditional lectures, these courses focus on reading and discussing papers from the literature. Like our other courses, CSC67XX courses involve multi-week, hands-on technical projects. Small groups meet with the instructor once per week to discuss the readings and projects. These courses require more out-of-class work than others in our program. To succeed, students should be comfortable working independently and possess strong time management skills.

CSC6712 Distributed Storage Systems

Students with a background in data structures, operating systems, or databases and a working knowledge of a systems language such as C, C++, or Java will be particularly well-prepared to succeed in this class. Those without this background may face an especially high level of challenge.

CSC7901 Machine Learning Capstone

Enrollment Steps

The steps for enrolling in the capstone project course are:

1. During discussions with your MSML advisor determine which semester you will do your capstone and receive approval from the capstone coordinator to begin your capstone in that semester.

2. By week 7 of the semester BEFORE you intend to do your capstone:
   1. Submit a paragraph describing your project idea to the capstone coordinator. Proposed projects must be well defined, feasible, and there has to be an available faculty member to supervise in order for them to be accepted.
   2. The capstone coordinator will evaluate your proposal to ensure that is is well-defined and feasible.
   3. The capstone coordinator will communicate any required revisions to the project or approval via email.
   4. Keep the capstone coordinator updated if your project idea changes significantly and on any advisor preferences.
3. By week 12 of the semester BEFORE you intend to do your capstone:
   1. The capstone coordinator will notify you who your advisor will be.
   2. Discuss and further refine your project idea in consultation with your advisor.
   3. Work with your capstone advisor to identify 2 other MSOE faculty members to serve on your committee. Outside committee members (e.g., the chief data scientist at your company) are allowed via adjunct appointments if they have appropriate experience and background, including at a minimum a relevant master’s degree.
   4. Complete the form (available here), get required signatures (from your advisor, from the capstone coordinator, and then from the CSSE Department Chair). The chair will submit it to the registrar and you will automatically be enrolled. See the form for the deadline, but submit it sooner if possible.

Capstone Project Options

Students may work:

• Individually,
• As part of a loosely affiliated group (e.g., regular lab meetings with advisor(s) for feedback on work in progress),
• As a team of two on a large project with clearly defined individual goals. Each student must submit their own capstone request form regardless of the project type.

Potential Capstone Project Advisors

Please contact these faculty and others to help refine your project idea:

• Dr. Berisha: Machine/deep learning, image processing, time-series data analysis. In general, my research centers around the application of machine learning algorithms in the fields of biomedicine and remote sensing. My most recent efforts have concentrated on two significant applications. Firstly, I have been exploring the integration of deep learning with hyperspectral mid-infrared imaging to enhance cancer diagnosis. Secondly, my research is dedicated to the application of deep learning techniques in the analysis of ECG data for the diagnosis of cardiovascular, brain injury, and respiratory diseases.
• Dr. Bukowy: Data Science, Experimental Design, Machine Learning, and Image Processing. Specifically, I have a lot of experience in medical image processing in bright field histology and MRI sequences for object detection.
• Dr. Durant: Signal processing, real-time systems, audio
• Dr. Flynn: Software Development, Software Process, Databases, Video Game Development. I’m interested in serving as a committee member on capstone projects, particularly when those projects involve aspects of software development or database integration.
• Dr. Imas: Medical imaging processing applications in neurology, cardiology and oncology, as well as electrophysiological applications of digital signal processing (e.g., EEG, ECG, EGM signals)
• Dr. Kedziora: Data science, statistics, machine learning, reinforcement learning
• Dr. Lukas: TinyML, Robotics, Autonomy, and embedded electronics
• Dr. Nowling: Data science, ML and data-intensive systems (production, recommendation, distributed, etc.)
• Dr. Urbain: Natural language processing, medical image processing, graph machine learning
• Dr. Wright: I am an astronomer with experience using machine learning for background rejection and object classification in optical images. I look forward to advising projects that involve astronomical data sets of any kind.
• Dr. Yoder: Large Language Models (LLMs, e.g., chat agents), Systems built on LLMs, Diffusion (image synthesis), theory of deep learning, Rosie-scale deep-learning experiments

Capstone Project Process

1. Committee Role: A capstone committee member is available as a technical resource throughout the project and participates in the midterm review and final presentation.
2. Weekly Meetings: The advisor and student will discuss project planning and intermediate results. The advisor may request a memo or slides to guide weekly discussions.
3. Initial Meeting: In the first week of the term, the student presents a preliminary project timeline, outlining tasks such as background research, addressing ethical and governance issues, data collection, data preprocessing and exploration, model exploration, implementation, preliminary results, model refinements, evaluation, final paper writing, and presentation preparation. Students may share drafts for input before this meeting.
4. Midterm Review: Early in the term, the student must contact the committee to schedule a 50–80 minute meeting at least one month prior to the final presentation to present preliminary results in detail and a revised plan for project completion.
5. Final Presentation: By mid-term, the student must contact the committee to schedule a 50–80 minute final presentation around the Monday of finals week. One week prior, the near-final report is due to the committee. The committee provides feedback before, during, or immediately after the final presentation. The student incorporates this feedback, including any changes requested, and submits an updated final report during finals week. The advisor will determine if further review from the committee is necessary.

Final Report

The final report length varies depending on the project content, but it is typically 25–40 pages, single-spaced, excluding any appendices. It is a formal report and should include a cover page and a table of contents. Typical sections include: Abstract, Introduction, Background (problem context and, optionally, research), Data (covering sources, acquisition, cleaning if applicable, and exploration), Methods (e.g., metrics used and general analysis methods), Baseline Model, Machine Learning Model, Discussion, Future Work, Conclusions, References, and Appendices.

The model sections often include many of the following subsections: Data Augmentation, Feature Engineering, Model Architecture, Hyperparameter Search, Results, and Discussion. Appendices can be used to delve deeper into architectures used, present detailed results that might interrupt the flow of the main paper, and discuss source code design and repository organization. Sections may be added or removed as appropriate to the project content.

Here are some examples of high-quality final reports.

PHL6001 AI Ethics and Governance

For MSOE undergraduates, this class cannot meet the undergraduate ethics requirement, and the undergraduate ethics requirement cannot meet this MSML requirement.

Many undergraduate majors allow a choice of ethics class. Students in these majors, if intending to pursue the MSML, should consider taking PHL3101 Ethics for Professional Managers and Engineers or PHL3103 Bioethics, which have almost no overlap with PHL6001. They may also take PHL3102 Ethics of Digital Technologies and AI, which has a partial overlap with PHL6001.

Course Details

Click on course numbers below to see catalog entries with course details.

• Students may not receive credit for CSC5610 if they have completed CSC2611 AI Tools and CSC2621 Intro to Data Science. Instead, they take an MSML elective.
• Offering information is fairly firm but might change.

Past and Planned Offerings

• Summer, 2023
  • CSC5120: Dr. Flynn
• Fall, 2023
  • CSC5610: Dr. Nowling and Dr. Bukowy
  • CSC5651: Dr. Durant
  • CSC6605: Dr. Nowling
• Spring, 2024
  • CSC5201: Dr. Urbain
  • CSC5601: Dr. Durant (in person)
  • CSC5610: Dr. Nowling (off-term offering)
  • CSC6621: Dr. Berisha (2 sections)
  • CSC6711: Dr. Nowling
• Summer, 2024
  • CSC5120: Dr. Flynn
  • CSC6621: Dr. Berisha (off-term offering)
  • CSC6712: Dr. Nowling (3 sections)
  • PHL6001: Dr. McAninch (2 sections, 7-week subterm 1)
• Fall, 2024
  • BME5210: Dr. Imas (2 sections, in person)
  • CSC5201: Dr. Urbain
  • CSC5601: Dr. Bukowy (in person)
  • CSC5610: Dr. Nowling (2 sections)
  • CSC5611: Dr. Yoder (in person)
  • CSC5661: Dr. Kedziora
  • CSC5980 An Engineering Approach to Game Theory and Economics: Dr. Thomas
  • CSC6711: Dr. Nowling (very limited capacity; closed section requests will be considered in August)
  • MTH5810: Dr. Armstrong
• Spring, 2025
  • BUS6141: Dr. Hrdlicka (asynchronous)
  • CSC5201: Prof. DeMastri (off-term offering, online)
  • CSC5611: Dr. Yoder, (in-person)
  • CSC5980 AI for Search: Dr. Urbain (online)
  • CSC6605: Dr. Nowling (2 sections, online)
  • CSC6621: Dr. Berisha (2 sections, online)
  • PHL6001: Dr. McAninch (online)
• Summer, 2025
  • BUS6141 Analytics Leadership and Strategy: Dr. Hrdlicka (May 19–July 5)
  • CSC5120 Soft Dev for ML: Dr. Flynn
  • CSC6621 Applied ML: Dr. Berisha
  • CSC6711 Rec Sys: Dr. Nowling (2 sections)
  • CSC6712 Distributed Storage Systems: Dr. Nowling (2 sections)
  • PHL6001: Dr. Talamentez (full term)
• Fall, 2025
  • BME5210: Dr. Imas (in person, evening)
  • CSC5201 Microservices and Cloud Computing: Dr. Urbain
  • CSC5601 Theory of ML: Dr. Bukowy (in person)
  • CSC5610 AI Tools and Paradigms: Dr. Nowling (1 section)
  • CSC5661 Reinforcement Learning: Dr. Kedziora
  • CSC5980 Survey of Deep Learning: Dr. Yoder (online, late afternoon)
  • CSC6605 ML Production Systems: Dr. Nowling
  • MTH5810: Dr. Armstrong