ISE 789-005/OR 791-005 Syllabus

ISE 789-005/OR 791-005 Syllabus

Spring 2024

Time

Thursday 3:00 P.M. – 5:45 P.M.

Room

02116 Textiles Complex

Instructor

Dr. S.-C. Fang (fang@ncsu.edu)

  • Office: 4341 Fitts-Woolard Hall
  • Office Hours: Tu 2:30 P.M. – 3:30 P.M. and Th 1:30 P.M. – 2:30 P.M. (or by appointment)

Teaching Assistant

Xinyu Xu (xxu37@ncsu.edu)

  • Office: 4333 Fitts-Woolard Hall
  • Office Hours: Wednesday 12:00 P.M. – 2:00 P.M. (or by appointment)

Prerequisite

  1. Linear Algebra and Linear Programming
  2. Ability to self-conduct computational experiments using platform solvers and programming languages such as MATLAB, Python, ILOG CPLEX, etc.

Course Objectives

This graduate-level course is an introductory course of supervised learning with the aim to introduce the basic concepts, models, methods, and applications of “Support Vector Machines (SVM)” and “Neural Networks (NN)” for machine learning. The course is designed for STEM students to

  1. understand the involved OR/Optimization models for classification, prediction, and generalization of machine learning techniques;
  2. experience the process of modeling, computing, and analytic decision-making in machine learning;
  3. work on some data banks with real applications.

Course Content

  1. Review
    1. Linear Algebra
      – Vectors and matrices, linear transformation, ranks,
      row/column/null spaces,
      eigenvalue/eigenvector & eigen-decomposition,
      singular value & singular value decomposition (SVD),
      principal component analysis (PCA)
    2. Optimization
      – Functions, gradients, unconstrained optimization,
      linear programming, quadratic programming
    3. Software Platform and Data
      – MATLAB, Python, Data banks
  2. Support Vector Machines
    1. Bi-classification
      – Linear SVM, soft-margin linear SVM,
      SVM with kernel, kernel-free SVM
    2. Multi-classification
      – OVA and OVO, Twin SVM
    3. Prediction
      – Support vector regression (SVR)
  3. Neural Networks
    1. Basic structure of NN
      – Neuron and activation function, perceptron,
      basic feed-forward neural network (NN)
    2. Backpropagation and learning
      – Error/Reward function, batch vs. online learning,
      learning algorithm
    3. Multi-layer neural network and deep learning
      – Scale, feature and computation, ReLU and SGD
    4. Radial basis function neural network (RBFN)
    5. Convolutional neural network (CNN)

Homework

  • Solution will be provided by recitation.
  • No late homework without pre-approval.

Project

  • End-of-semester report and presentation.
  • At most 2 persons a team.

Exam

Final project presentation will be scheduled for oral exam.

Grades

  • Grading of the course is mainly homework-project based.
  • Homework assignment with each lecture (60%).
  • Final project (40%).

Evaluation Standard

  • A -88 and above
  • B -75 to 87
  • C -60 to 74
  • Fail -under 60

Classroom Rules

  • Rule 1: No late homework without pre-approval.
  • Rule 2: Turn in your homework through email.
  • Rule 3: Office hours can be held through Google Meet.

End-of-Semester Class Evaluation

Online class evaluations will be available for students to complete during the last two weeks of class. Students will receive an email message directing them to a website where they can log in using their Unity ID and complete evaluations. All evaluations are confidential; instructors will never know how anyone student responded to any question, and students will never know the ratings for any particular instructors.

Evaluation Website: https://classeval.ncsu.edu

Student help desk: classeval@ncsu.edu

More information about ClassEval: http://www2.acs.ncsu.edu/UPA/classeval/index.htm

Academic Integrity

A student is expected to know what constitutes academic misconduct found in the Code of Student Conduct Policy ( POL11.35.1), and to take responsibility for his/her actions. A student who is unsure whether an action constitutes an offense, or who needs help in learning how to avoid offenses (e.g., plagiarism, cheating) or about “rules” for group work/collaboration should seek guidance from the course instructor.

Students with Disabilities

North Carolina State University retains authority, through the Disability Services Office (located in Student Health Services Building, Suite 2221), in determining appropriate accommodations after giving consideration to the preferences of the student, the documentation provided, and institutional expertise in working with students with disabilities. If you require academic accommodations to lessen the impact of your disability, please register with the Disability Services Office at the beginning of each academic term.