6. Term Project

• Project weeks: March 23 ~ April 5 (two weeks)
• It is an individual, NOT team, project. You need to decide the term project yourself.
• Requirements - part I
  • Your project should NOT be shared for any other courses.
  • Your project SHOULD be approved by the instructor after you submit your project proposal. You may need to submit the project proposal early so that your proposal can be approved early.
  • During the project weeks, you SHOULD meet the instructor ONCE a week to discuss about your project. Technical problems, progress, direction, etc will be discussed. Project meetings will be held only on Monday and Tuesday.
• Requirements - part II
  • Your project should use JavaScript. When you decide to use another programming language, you need to get the approval first.
  • Your project should NOT use any code from the Internet.
  • Your project should NOT use any library unless the use of the library is approved. When you decide to use any library or tool published in the Internet, you need to have a discussion with the instructor to measure how much beneficial it would be to use the library/tool.
  • Your project should be the implementation of a FULL application. Nothing should be hard coded, if possible. E.g., number of hidden layers and number of nodes in each hidden layer in a neural network.
• Some example projects are
  • A* algorithm - Navigation system for BC
  • A* algorithm - Rubik's cube
  • A* algorithm - Path finding in a terrain having obstacles
  • Local search - TSP
  • CSP - The n-queens problem with the least constraining variable first heuristic, or with the most constrained variable first heuristic
  • Genetic algorithm - The n-queens problem
  • CSP or genetic algorithm - Class scheduling problem
  • Expert system - forward chaining for the ZooKeeper problem
  • Expert system - resolution refutation for the ZooKeeper problem
  • Decision tree - Box-office like data set
  • Clustering - K-Means algorithm for data of numerical attributes
  • k-Nearest neighbor algorithm with reconfigurable k and multiple ks at the same time
  • Neural network - Backward propagation with delta rule for alphabet recognition, with the non-alphabet case
  • Neural network - Hopfield or BAM for alphabet recognition, with the non-alphabet case
  • Neural network - SOM - 3D numerical data into 2D, with visualization
  • Bayesian belief netowrk - Burglary, or something else
  • Naive Bayes classifier
  • Fuzzy control system - Toy 2-wheel vehicle
  • ... your own project idea ...
• Submissions
  • Project description and design document. Here is a sample.
  • Source code of your application - NO zipped file because the TRU email system blocks any zip file.
  • Data set if applicable
• Due
  • Submission of the project proposal: March 25, 2019
  • Submission of the 1^st report including source code: March 29, 2019
  • Submission of the final report including source code: April 5, 2019
• Evaluation
  • The project proposal - 10%; Here is a sample.
  • Implementation - 90%
    • Completeness of the 1^st submission - 30% (The half of the whole implementation is expected to be completed.)
    • Completeness of the final submission - 60%
  • Final grade = beta * (Documents + Implementation), where beta (∈ [0,1]) is the level of workload of your project.
  • 10% bonus when you have a presentation in the CS Showcase. You need to get the approval.
  • Sorry, but ZERO for any of the following cases.
    • Any project that did not have 2 project meetings
    • Any late submission
    • Any submission of unapproved project. It will be considered as a copy work.
    • Any copy work. There have been some cases of this. Academic misbehavior will be reported according to the TRU policy.
    • Any hand-written documents
    • Any final submission that includes any syntax errors