Requirements

This section summarizes requirements for students working on the projects of 3C group. If you have any question regarding the requirements, please ask your supervisor.

  • The student needs to regularly meet with the supervisor once a week.
  • Each meeting will be recorded in the Progress file (which resides in Google Drive directory - see below).

  • The record will contain:
    • The date of meeting.
    • Review of the results of the last week tasks.

    • List of tasks for the next week. The tasks are small enough to complete them in one week.

  • Student will work on the project at least 6 hours a week in the LAB KN:E 224. They will share the progress with other students.

  • Student will provide picture and small CV for the people pages on the 3c web pages.

  • All documents and reports will be in English. All documents will be located in Google Drive directory with a name FIRSTNAME_LASTNAME  shared with supervisor. In case we need to refer to externally stored data, the links to other documents need to be recorded in the Progress file.

  • The source code, if not agreed differently, will be stored in Github, The link needs to be recorded in the Progress file.

The Research project milestones

  • Essay - max 10 sentences long recorded in a Google Drive students directory with a name Essay_FISTNAME_LASTNAME. This document should explain why we try to solve the problem in general words, what would be the benefits and who needs it.

  • State of the art analysis - a document of about 10 pp. size containing the following items:

    • Description of the problem - short but with more technical details than the essay

    • Top 10 most influential papers in the last 5 years, with a short description. Hint: find the key words and use Google scholar, focus on articles with top number of references.

    • List of  top 3 experts (names) and teams (location) working on the problem. Derive from the previous step

    • List at least 2 of the top journals and an 2 conferences about the topic - derive them from where the papers were published

    • List open competition in the field, if they exist

    • Indicate the first steps - for example which of the algorithmus from the most influential articles we should try first to set the initial baseline error rate.

    • Plan the first steps - record the first questions you need to answer to complete the previous milestone. i.e. where do I find a sample code, where do I get a training/testing data, what is the quality measure, what computers do I need?

    • Suggest how will be the experiments recorded, i.e. what all needs to be recorded to give sufficient information for follow up work.