Queen's School of Computing

Practical applications of AI have exploded in recent years with a resurgence of research in machine learning and, especially, in deep learning. Real-world applications include medical diagnosis, text and voice analysis for big data applications, computer vision, and self-driving cars.

There is a growing demand for graduates with an AI background from the leading technology firms, healthcare companies, automobile manufacturers, and research labs. The creation of a "Field of Study in Artificial Intelligence" will help to distinguish Computing graduates and increase their competitiveness for these highly skilled positions.

The School of Computing has substantial expertise in this Field, with core faculty members in AI of Farhana Zulkernine (specializing in neural networks), David Skillicorn (specializing in data mining), Dorothea Blostein (specializing in pattern recognition), and Randy Ellis (specializing in nonlinear optimization).

The requirements of the Field are that the student take three courses in AI (described below), complete an AI-related thesis, and complete the other requirements of the degree program. Graduates that complete the requirements will recieve the "Field of Artificial Intelligence" designation on their transcript.

MSc Program

For the MSc (research-based only), the Field of Study in Artificial Intelligence will require that student:

  • take three of the six AI courses listed further below, including at least one of CISC 856 and CISC 867.
  • take two more courses as required in the MSc program; and
  • complete an AI-related thesis.

PhD Program

For the PhD, the Field of Study in Artificial Intelligence will require that student:

  • satisfy three tokens that show graduate-course-equivalent experience in three of the six AI courses listed further below, including at least one of CISC 856 and CISC 867.
  • satisfy seven more tokens as required in the PhD program; and
  • complete an AI-related thesis.

Courses

  • CISC 856*/3.0 Reinforcement Learning
    This course includes topics on formal and heuristic approaches to problem solving, planning, reinforcement learning, knowledge representation and reasoning, Markov decision processes, dynamic programming, temporal difference learning, Monte Carlo reinforcement learning methods, function approximation methods, integration of learning and planning.
    Prerequisite: CISC-352 or equivalent, programming expertise.
  • CISC 867*/3.0 Deep Learning
    Teaches algorithms and concepts about deep learning based on the biological neural network. Students will learn about deep belief network, restricted Boltzmann machine, Convolutional, Generative adversarial and Long Short Term Recursive NN and develop DNN using tools such as TensorFlow to perform feature extraction, image recognition and text processing.
    Prerequisite: COGS 400/ CISC 452/CMPE 452 or co-requisite CISC 874
  • CISC 855*/3.0 Nonlinear Optimization
    Methods for computational optimization, particularly examining nonlinear functions of vectors. Topics may include: unconstrained optimization; first-order methods; second-order methods; convex problems; equality constraints; inequality constraints; applications in machine learning.
  • CISC 859*/3.0 Pattern Recognition
    An introduction to statistical and structural pattern recognition. Feature extraction and the feature space. Bayes decision theory. Parametric classification. Clustering methods. Syntactic pattern description: string, tree and graph grammers; attributed grammars; stochastic grammars. Error correcting parsing; parsing of stochastic languages. Assignments include practical experience in application areas such as character recognition and document image analysis. Three term-hours; lectures and seminars.
  • CISC 873*/3.0 Data Mining
    Study of the extraction of concepts from large high-dimensional datasets. Statistical foundations; techniques such as supervised neural networks, unsupervised neural networks, decision trees, association rules, Bayesian classifiers, inductive logic programming, genetic algorithms, singular value decomposition, hierarchical clustering.
    Three term hours; lectures and seminars.
  • CISC 874*/3.0 Neural Networks
    An introduction of the computational properties of neural networks. Topics may include: Learning Processes, Single Layer Perceptrons, Multilayer Perceptrons, Principal Components Analysis, and Self-Organizing Maps.
    Three term hours; lectures and seminars.