Caution: This course is expected to evolve throughout the semester!

Refinements are likely.

---

Overview

Instructor

Office Hours

Some Possible Topics

Course Format

Academic Honesty

Electronic Media

Accommodations

Grading

Algorithms are designed to address specific needs, but these algorithms also may have unintended consequences. For example, in her book, Weapons of Math Description (Crown Publishing, 2016), Cathy O'Neil describes "how big data increases inequality and threatens democracy." This tutorial will examine approximately five application areas. Discussion of each topic will include

• problems being addressed,
• selected details of the underlying algorithmic solutions,
• effectiveness of the solutions, and
• potential risks.

Initial work for this tutorial will use O'Neil's book as a starting point.

• what data are collected for several socially-oriented applications,
• what algorithms are used,
• to what extent are results biased,
• if the work is biased, what factors contribute to that bias, and
• how might one measure that bias?
• what approach(es) might be used to remove or mitigate any bias?

The selection of later topics will depend upon the interests of students in the course.

For each such application, students will first determine a [partial] list of technical, social, and ethical issues. Attention will then focus on needed hardware, networks, algorithms, and other implementation challenges. With this background, students will explore the current state-of-the-art for modern systems. Potential topics for exploration include

• what results are achieved successfully,
• how might systems be tested,
• what risks might arise,
• what impact the systems might on various stakeholders, and
• how might challenges be addressed?

Henry M. Walker

Office: TCL 102
Telephone: (413) 597-5045
E-mail: hmw3@williams.edu

• Monday, Wednesday: 3:00–4:30 pm
• Tuesday, Thursday: 10:00–11:30 am
• Friday: 1:00–2:30 pm

Adjustments to this schedule will be posted on the board outside my office.

Beyond the initial consideration of Big Data, its assumptions, algorithms, and uses, students and the instructor will collaborate in choosing 3–4 additional subjects for exploration. Some promising topics include

• Amazon's cashier-free grocery store (reportedly needing only 6 employees)
  • Possible initial sources:
    • Company announcement: amazon.com/go
    • News article in GeekWire.com: Amazon to open first self-driving grocery store, with no checkout lines, in Seattle in early 2017
    • Modest commentary in GeekWire.com: The end of grocery checkers? Amazons high-tech store points to the future of physical retail

• Electronic voting machines

  • Possible initial sources:
    • Erica Naone, "Voting with (Little) Confidence", MIT Technology Review, https://www.technologyreview.com/s/409453/voting-with-little-confidence/, 29 January 2008 (Accessed 10 August 2017).
    • Sheila Parks, While We Still Have Time: The Perils Of Electronic Voting Machines And Democracy's Solution: Publicly Observed, Secure Hand-Counted Paper Ballots (HCPB) Elections, Apparently with some connection to the Center for Hand-Counted Paper Ballots, 6 September 2012.
    • Mike Orcutt, "A Close Election Could Expose Risky Electronic Voting Machines", MIT Technology Review, https://www.technologyreview.com/s/602482/a-close-election-could-expose-risky-electronic-voting-machines/ 30 September 2016 (Accessed 10 August 2017).

• Algorithms within autonomous vehicles

  • Possible initial sources:
    • Andrea Peterson, "Google didn't lead the self-driving vehicle revolution. John Deere did.", The Washington Post, https://www.washingtonpost.com/news/the-switch/wp/2015/06/22/google-didnt-lead-the-self-driving-vehicle-revolution-john-deere-did/, 22 June 2015 (Accessed 10 August 2017).
    • Frank Tobe, "As Google cars roll out, Deere reminds us of thousands of their self-driving tractors", The Robot Report, https://www.therobotreport.com/news/as-google-cars-roll-out-deere-reminds-us-that-they-have-thousands-of-self-d, 27 June 2015 (Accessed 10 August 2017).
    • NASA, "How NASA and John Deere Helped Tractors Drive Themselves", https://www.nasa.gov/feature/directorates/spacetech/spinoff/john_deere, 16 November 2016 (Accessed 10 August 2017).
    • Jamie Condliffe, "Are Autonomous Cars Ready to Go It Alone?", MIT Technology Review, https://www.technologyreview.com/s/603883/are-autonomous-cars-ready-to-go-it-alone/, 17 March 2017 (Accessed 10 August 2017).

• Google's PageRank Algorithm

  • Possible initial source:
    • Amy N. Langville and Carl D. Meyer, Google's PageRank and Beyond: The Science of Search Engine Rankings Princeton University Press, 2006.

• Bitcoin and its Use in Commerce

  • Possible initial source:
    • Arvind Narayanan, Joseph Bonneau, Edward Felten, Andrew Miller, and Steven Goldfeder, Bitcoin and Cryptocurrency Technologies: A Comprehensive Introduction, Princeton University Press, 2016.

• The P = NP Question and its Far-reaching Implications

  • Possible initial source:
    • Lance Fortnow, The Golden Ticket: P, NP, and the Search for the Impossible, Princeton University Press, 2013.

• Electronic funds transfer systems

  • Possible initial sources:
    • Technology Review,"40 Years Ago: Electronic Money Is Too Easy" Technology Review, https://www.technologyreview.com/s/603606/40-years-ago-electronic-money-is-too-easy/, 22 February 2017 (Accessed 10 August 2017).
    • Online Q/&A, "Is this legit, or should I have a fit?", Ask MetaFilter, http://ask.metafilter.com/159439/Is-this-legit-or-should-I-have-a-fit, 13-14 July 2010 (Accessed 10 August 2017).

• Hospital Monitoring and Response

• Airline Reservation Systems

In addition, the following sources may provide helpful background.

• Sara Baase and Timothy M. Henry, A Gift of Fire: Social, Legal, and Ethical Issues for Computing Technology, Fifth Edition, Pearson Education, 2018.

• Panos Louridas, Real-World Algorithms: A Beginner's Guide", MIT Press, 2017.

• Cindy Meyer Hanchey, "Computer Ethics Links", Special Interest Group in Computer and Society , http://www.sigcas.org/ethics (Accessed 10 August 2017). (Some links may be out of date.)

Background

The Williams College Catalog states:

Tutorials place much greater weight than do regular courses—or even small seminars—on student participation. They aim to teach students how to develop and present arguments; listen carefully, and then refine their positions in the context of a challenging discussion; and respond quickly and cogently to critiques of their work. Tutorials place particular emphasis on developing analytical skills, writing abilities, and the talents of engaging in rigorous conversation and oral debate.

The Williams home page for Tutorials highlights interactions of two students with one profess for "An in-depth conversation, fueled by intellectual curiosity and the spirit of debate, that takes place over the course of an entire semester", following "the Oxford University style of education."

An article on Williams' tutorials in the Chronicle of Higher Education highlights at least two different models for the format of tutorials:

• Students are organized into pairs, each working on an individual project. In weekly meetings one students presents the current draft of a paper or research project, and the other student provides feedback and constructive criticism—all under the guidance of a faculty member.

• In a physics "tutorial on electromagnetic theory, the students don't offer papers but take turns at the chalkboard, going over the answers to a lengthy problem set."

Format for this Course

This tutorial combines elements of several instructional models, while maintaining the themes of student engagement, close student-faculty interaction, sharpening technical and analytical thinking, and exploration of substantial topics and issues.

Although details will evolve throughout the semester, key elements of the course format are envisioned as follows:

• The tutorial will organize the semester into 4-5 segments.

• Each segment will focus on one topic, with exploration extending over approximately 3 weeks.

• For each segment and topic, each student will identify one or more goals, issues, and products.

  • During the first week of each segment, each student will prepare a list of issues that the student will investigate.
  • Products for a segment might include a paper, problem set, poster, computer program, mathematical analysis, presentation, or other tangible result.
  • To gain practice in professional writing,
    • products for at least two segments must include an article, written in the form and style required for publication in ACM Inroads, the computer science education magazine for the Association for Computing Machinery (ACM).
    • instructions and information for Inroads authors may be found through a link from the submission home page at https://mc.manuscriptcentral.com/inroads.
  • After the first week of a segment, students will delve into the topic, with the intent of preparing their products.

• All members of the tutorial will meet together once per week to reflect on previous work, determine weekly milestones, and clarify forthcoming activities. These whole-group sessions will extend about one hour.

• In working on a topic, the class will be divided into subgroups of two (or occasionally three, if an odd number of students enroll in the course). These subgroups may change from topic to topic through the semester.

• Each subgroup will meet with the instructor individually for about an hour each week, focusing on the goals and products identified for that topic.

• Although initial work will develop products for local use (e.g., in-class presentation and peer review), periodic discussion can be expected regarding the possible submission of selected products to a regional or national student poster event, paper session, or other external venue.

Work Load

Tutorials at Williams require substantial student initiative, effort, thought, and follow-through, as well as substantial student-faculty interaction.

• Students help identify topics and set directions, explore and research topics, analyze materials and determine conclusions, and structure ideas into presentations and reports.

• The instructor provides guidance, mentoring, and feedback.

Such work requires time for both students and the instructor, and students should expect to devote 12-15 hours per week to this course, in addition to the weekly group and collaborative-pair sessions.

All work in this course is governed by Williams' Honor Code and the rules of the college regarding academic honesty. In summary, standard practice requires that you must acknowledge all ideas from others.

• When working either individually or collaboratively, you must cite all statements from any written source, if you use them to guide your work.

• For any work, normal rules apply for quotation, paraphrase, citation, and bibliographic reference sections.

• Much work in this course will be done individually, but some activities may be collaborative. When collaborating, the names of students in the group should appear as authors. Further, the listing of several authors implies that all members of the group agree with what is presented. If a group member does not agree with some part of the work, the group should continue to discuss and revise the material until agreement is achieved. In summary, a group activity is a joint effort, and all group members have equal responsibility for the finished product.

• When you work on an activity yourself, but consult others, then you should include a statement identifying whom you consulted on what material. This includes conversations with other students, faculty, and any other people involved. If you consulted one person on several parts on an activity, you may summarize the collective help in an "Acknowledgment Section" rather on individual sections or subsections as long as the work makes clear who helped on what. Overall, all consultations on all work must be cited for each problem, exercise, article, poster, etc.

Finally, please note:

• Although the Web can be useful for reference, you are advised that much material on the Web is of poor quality.

• You are responsible for the quality of what you turn in, regardless of the source of the material.

If you have any questions about how the honor code applies to your work, please come talk with me. I am always happy to have those conversations.

Cell phones, text-messaging devices, and other social-networking connections may not be used during class sessions. If you bring such equipment to the classroom, it must be turned off before the class starts and stay off throughout the class period. Use of such equipment is distracting to those nearby and will not be tolerated.

Students with disabilities of any kind who may need accommodations for this course are encouraged to contact Dr. G. L. Wallace (Director of Accessible Education) at 597-4672. Also, students experiencing mental or physical health challenges that are significantly affecting their academic work or well-being are encouraged to contact me (as course instructor) or to speak with a dean. The deans can be reached at 597-4171.

This instructor's grading philosophy dictates that the final grade should ultimately be based upon the depth, scope, and quality of each student's work, not on the performance of the class as a whole nor on a strict percentile basis.

According to the Williams' Registrar's Office, letter grades at Williams have the following meaning:

• A = excellent;
• B = good;
• C = fair;
• D = passing;
• E = failing.

Grading in this course will take these descriptors seriously.

---

created 8 August 2017 revised 8-20 August 2017
For more information, please contact Henry M. Walker at hmw3@williams.edu.