Connor's Notes

⬅️ Back to portfolio

COSC 3P71: Tutorial 6

3 min read

Report Writing (assignment 2)

  • doesn’t need to be perfect → don’t worry if tables are too wide etc.

Abstract

  • summarize the content of the report
  • should be simplified & self contained (avg. cs student should be able to read and understand it without any outside reference)
  • no abbreviations, equations, tech specs, …
  • ~150-250 words
  • sections
    • introduction
      • broad → narrow
    • methodology
      • leave out numbers !!!
      • “tested various GAs with multiple setups” → does not need to be in depth
    • results
      • no numbers !! 💥
      • nice and simple, ~1-2 sentences

Introduction

  • “freest” location to put citations
  • introduce the general topic being covered (high level description, how the work applies to the real world, …)
  • discuss the problem & approach to solving the problem
    • give more detail than in the abstract
    • explain what the problem is, why does it need solved?
    • how are you going to solve the problem? is your approach different than other/naive approaches?

not necessary for this assignment

Background

  • give the reader information they may not know that is required to understand your report
  • define algorithmically, how many algorithms you used for this assignment
    • pseudocode for a loop for a GA
  • expect the reader to have a general understanding of the field, but not the details of your problem & methodology (assume they are a cs student that has not taken this course)
  • consider importance of the topic for the report (if it’s very important → should probably add a subsection describing it)
  • not really required for this assignment

Proposed approach

  • introduced a new approach for solving the problem
  • “set yourself apart”

Experiment Setup 💥💥💥

  • VERY IMPORTANT
  • explain what experiments you will perform
  • how will performance be measured?
  • statistical tests?
  • give all parameters used for the experiments & statistical tests
  • include crossover, selections, … explanation with pseudocode if possible
  • include pop size, generations, etc
  • report if custom seeds (for randomizations) were used & what they are

Results

  • provide results of experiments
  • create tables, graphs, figures, where appropriate
    • remember appropriately labeled axis of your graphs
  • try to make your results easy for reader to understand
    • bold font, highlight, italics, visuals (figures) ← for FIGURES, kinda weird for regular text
  • provide a brief discussion of results
    • describe what is being shown in figures/tables/graphs
    • which approach performed best?
  • reference all tables, figures etc. no matter what

Conclusion

  • summarize the report
  • essentially the abstract except with more emphasis on the results
    • high level discussion about the most meaningful results
    • interpret the results to give “big picture” ideas
  • any shortcomings of the report, limitations, issues you couldn’t resolve
  • can also discuss ideas for future work

References

  • APA format preferable (according to slides)
    • just use block number for intext citation
  • refer to the references in-text when appropriate
  • cite the slides ← could be considered academic integrity…?
  • use up to date references → slides are up to date → freelo

General Writing Guidelines

  • use formal language (no “cuz”, “u”, etc.)
    • DO NOT DO what i do in my notes lol
  • conclusions should be backed up by results, refer to them in-text when appropirate
    • e.g. “As shown in Table 3, Alpha-Beta pruning was 70% faster than Minimax”
  • claims that are not general knowledge should be supported by a reference
    • genetic algorithms IS NOT GENERAL KNOWLEDGE, cite it

Extra

  • 3-5 pages is good
  • if there are a bunch of different line graphs, combine them

Graph View