CS 344 - Project

cs344 → project

The main lesson of thirty-five years of AI research is that the hard problems are easy and the easy problems are hard. The mental abilities of a four-year-old that we take for granted — recognizing a face, lifting a pencil, walking across a room, answering a question — in fact solve some of the hardest engineering problems ever conceived …. As the new generation of intelligent devices appears, it will be the stock analysts and petrochemical engineers and parole board members who are in danger of being replaced by machines. The gardeners, receptionists, and cooks are secure in their jobs for decades to come. — S. Pinker, The Language Instinct, 1994.

This course requires a substantive final project.

Requirements

Your final project deliverable must:

include a report discussing the: application domain; relevant theory on which the system based; reasons for choosing this theoretical foundation rather than others; broader implications of the application and its use.
include a substantive, implemented system that:
- addresses some applied area of AI, e.g.: Game theory and adversarial games; Image processing and vision; Robotics; Natural language processing; Information retrieval and extraction.
- uses one or more of the basic AI technologies, e.g.: Symbolic systems; Genetic algorithms; Neural networks; Statistical methods.

We reserve the right to limit the number of applications proposed for each project area or technology.

Timeline

There will be four “deliverables” for this project, to be submitted in your course repo in the directory project and due according to the following schedule:

Proposal (two drafts, 5% each) — Propose a project and develop the idea.
- Draft — The preliminary draft will be a two-to-three-paragraph vision statement discussing your domain and your proposed basic technology.
- Update — The final draft will be a one-to-two-page enhancement of your preliminary vision statement that includes preliminary drafts of all the sections required in the final paper.
Submit this as a Jupyter notebook (proposal.ipynb).
Walkthrough (10%) — Meet with us one-on-one to talk through your domain, theory and prototype. Be prepared to discuss the key elements of your work, your poster, and to give a demo of your system. We’ll identify the key work you still need to do and prepare you for the final showcase.
Showcase (10%) — Attend the final project showcase in which you can see everyone’s project. Be prepared to demo your system and to explain its key points using a Jupyter notebook (either a separate showcase.ipynb or your final report.ipynb, see below).
Submission (70%) — Submit your final project deliverable, including:
- the full code base (raw code modules); include a README.md file in the project root that specifies the name of the project, gives a brief overview of the vision of the project, lists the various code modules and their purpose, explains how to run the code, and references external sources for any required datasets. Remember, do not submit large dataset files in your repo; store them someplace else and include links to them in this README file.
- a brief summary of the upgrades you’ve made since the project walkthrough (updates.ipynb).
- a detailed report on your system and your results (report.ipynb). This should be on the order of 2000-2500 words (around four pages on “standard” paper, including bibliography but not including program code or raw output) and include material appropriate for your project. What is appropriate may vary from project to project, but would likely include the following.
  - Vision — Give an overview of the project and its purpose.
  - Background — Describe the technologies you’re using and why you chose them. Include citations of the work on which you’ve based your system. Include citations for the technologies we’ve covered in class (no description is required for these) and the new technologies you’ve experimented with (include descriptions of these).
  - Implementation — Summarize your implementation and, if appropriate, how it extends the work you’ve referenced.
  - Results — Give the results of your system and comparing them with other similar work.
  - Implications — Discuss the social and ethical implications of using the technologies you’ve chosen for your project.

See the policies page for the project due-dates and times. The points for these deliverables will be combined to compute your score for the final project.