CS 600 schedule

Spring 2017 Schedule of Topics

Jump to week[n] ==> 1, 2, 3, 4, 5, 6, 7, 8

2017-03-09 Week 8

• class not held, reading assigned.

2017-03-02 Week 7

GROUP A: Research Tools

• citing papers in text (Google Scholar ==> Zotero ==> Document)

2017-02-21 Week 6

GROUP A: Research Tools

Topics

• services for file and document management
  • Google Drive, Dropbox, Your own Github site (like this one), blog, website, etc…
• Zotero and citation management

Tasks

1. browse the Social Media Research Toolkit

2. find a tool that interests you

3. identify an academic paper associated with that tool (they’re linked next to the tool)

4. read the paper.

5. write a summary of the paper (.5 page at most), then use zotero to … a. include an inline citation b. include a bibliography section containing the full reference to the paper.

6. store your summary in your google drive or other online accessible location.

2017-02-14 Week 5

Data Science Topics, Projects, and Resources

Topics

• Graduate students organized into three groups (A, B, C)

2017-02-07 Week 4

Research Paper Discussions

see homework 1 for papers and day on which you will lead class discussion.

2017-01-31 Week 3

Research Papers/Videos

Security

• Where is cybercrime really coming from? (video)

the economy or “marketplace” for computer security criminals

NEW Approach: change the economics for the bad guys?

** published 700 terabytes of actionable threat intelligence data, including information on real-time attacks that can be used to stop cybercrime in its tracks. And to date, over 4,000 organizations are leveraging this data, including half of the Fortune 100.**

Machine Learning

• ‘Weapons Of Math Destruction’ Outlines Dangers Of Relying On Data Analytics (audio)

So in some sense, it’s really not the algorithm’s fault at all. It’s, in a large way, the way we apply algorithms and the way we trust them that is the problem.

people just trust numbers, they trust scores

worrisome algorithms:

1. high impact
2. opaque
3. destructive

• Accountability in Machine Learning (pdf)

In the face of important or expensive errors, discrimination, unfair denials of public services, or censorship, when and how should algorithms be reined in?

prioritize, classify, associate, and filter

One issue with the church of big data is its overriding faith in correlation as king. “Correlation does not equal causation”

algorithms are largely unregulated now, and they are indeed exercising power over individuals or policies in a way that in some cases (for example, hidden government watch lists) lacks any accountability whatsoever. recidivism???

Freedom of Information Processing Act

• What do we do?
  • transparency
  • allow adjustments to false positives
  • publicly submitted benchmark datasets
• What to disclose in the spirit of transparency?

human involvement, data, the model, inferencing, and algorithmic presence.

• Who’s running the algorithm? ==> attribution
• What is the data, is it accurate, valid, relevant, etc.?
• How is algorithm modeled? Eg., How does the Growth Model work?
• Inference testing. Use multiple and diverse datasets to benchmark an algorithm’s results.
• Transparency: algorithmic presence. What’s it watching and what is not watching…

machine to human explainers?

Topics From the Readings/Video:

• growth model
• predictive policing
• recidivism risk scoring
• automated writing algorithms
• algorithm transparency
• gaming the algorithm
• cybersecurity as health/disease pandemics
• threat intelligenc: how is it used, processed, made actionable?

Topics

• reading/video discussions

2017-01-24 Week 2

Research/Topic Presentations

Topics

• Tuesday ==> Year 1+ grad students present their thesis topic and progress.
• Thursday ==> First semester grad students present their research interests and areas for topic discovery.

2017-01-19 Week 1

Introduction

Topics

• course introduction
• syllabus