|
COSC 4P76 Machine Learning
|
Project Details:
Topic
The project is in a topic on machine learning, the topic must be approved by me beforehand.
You may propose your own topic related to machine learning (or choose one of my proposed topics, discussed in class). I will give details regarding choice of topic in class.
NOTE: You may work in pairs (and in veryspecial cases depending on scope
of project I may allow larger groups, but I rarely do). Of course you can also work alone, but I recommend working
in pairs as much as possible.
Project topic proposal (Due date: March 12th, 2018)
The purpose of the project proposal is:
- to get you started on your project;
- to identify a topic for your project;
- to define the extent and the scope of your project; and
- to receive instructors approval and feedback on the above items.
The project proposal must be at most 1 page long and include the following items:
- One paragraph that describes the project and the learning task in plain English
- A Precise description of the learning task identifying the task, performance measure, and training experience.
- Suggested choices for designing the learning system in terms of type and representation of training examples, type of neural networks, and the learning algorithm.
- Suggested experimental setup and analysis of results
- References to related material (publications, tools, web sites)
Project Report & Supporting code (Due April 20th, 2018)
The final project report (with supporting software) is due in April and is worth 25%
of your final mark. Late submissions will not be accepted.
Requirements and guidelines (See further detailed guidelines handed-out on day one):
- The report should be about 8 pages long including a reference page (plus one page excess is acceptable, any extra will not be graded) including references but excluding appendices.
- You must use an IEEE article template, no other format accepted.
- Raw data, source code, and the user manual should be included as appendices.
- Use of LaTeX for preparing the report is highly recommended.
- The report should include the following section, as appropriate:
- Abstract
- Introduction
- Problem definition, Background/Related Work
- Learning Algorithm & Methodology
- Experiments
- Analysis
- Conclusion and Future Work
- References
- You should include a number of references
- You must follow the department Assignment Guidelines when submitting your report.
|