beast-cancer-wisconsin dataset (Nearest-Neighbor Machine Learning Bakeoff)

Brief information from the UC Irvine Machine Learning Repository:

  • Donated by Olvi Mangasarian
  • Located in breast-cancer-wisconsin sub-directory, filenames root: breast-cancer-wisconsin
  • Currently contains 699 instances
  • 2 classes (malignant and benign)
  • 9 integer-valued attributes
  • Ftp Access

Citation Request:
   This breast cancer databases was obtained from the University of Wisconsin
   Hospitals, Madison from Dr. William H. Wolberg.  If you publish results
   when using this database, then please include this information in your
   acknowledgements.  Also, please cite one or more of:

   1. O. L. Mangasarian and W. H. Wolberg: "Cancer diagnosis via linear 
      programming", SIAM News, Volume 23, Number 5, September 1990, pp 1 & 18.

   2. William H. Wolberg and O.L. Mangasarian: "Multisurface method of 
      pattern separation for medical diagnosis applied to breast cytology", 
      Proceedings of the National Academy of Sciences, U.S.A., Volume 87, 
      December 1990, pp 9193-9196.

   3. O. L. Mangasarian, R. Setiono, and W.H. Wolberg: "Pattern recognition 
      via linear programming: Theory and application to medical diagnosis", 
      in: "Large-scale numerical optimization", Thomas F. Coleman and Yuying
      Li, editors, SIAM Publications, Philadelphia 1990, pp 22-30.

   4. K. P. Bennett & O. L. Mangasarian: "Robust linear programming 
      discrimination of two linearly inseparable sets", Optimization Methods
      and Software 1, 1992, 23-34 (Gordon & Breach Science Publishers).

1. Title: Wisconsin Breast Cancer Database (January 8, 1991)

2. Sources:
   -- Dr. WIlliam H. Wolberg (physician)
      University of Wisconsin Hospitals
      Madison, Wisconsin
   -- Donor: Olvi Mangasarian (
      Received by David W. Aha (
   -- Date: 15 July 1992

3. Past Usage:

   Attributes 2 through 10 have been used to represent instances.
   Each instance has one of 2 possible classes: benign or malignant.

   1. Wolberg,~W.~H., \& Mangasarian,~O.~L. (1990). Multisurface method of 
      pattern separation for medical diagnosis applied to breast cytology. In
      {\it Proceedings of the National Academy of Sciences}, {\it 87},
      -- Size of data set: only 369 instances (at that point in time)
      -- Collected classification results: 1 trial only
      -- Two pairs of parallel hyperplanes were found to be consistent with
         50% of the data
         -- Accuracy on remaining 50% of dataset: 93.5%
      -- Three pairs of parallel hyperplanes were found to be consistent with
         67% of data
         -- Accuracy on remaining 33% of dataset: 95.9%

   2. Zhang,~J. (1992). Selecting typical instances in instance-based
      learning.  In {\it Proceedings of the Ninth International Machine
      Learning Conference} (pp. 470--479).  Aberdeen, Scotland: Morgan
      -- Size of data set: only 369 instances (at that point in time)
      -- Applied 4 instance-based learning algorithms 
      -- Collected classification results averaged over 10 trials
      -- Best accuracy result: 
         -- 1-nearest neighbor: 93.7%
         -- trained on 200 instances, tested on the other 169
      -- Also of interest:
         -- Using only typical instances: 92.2% (storing only 23.1 instances)
         -- trained on 200 instances, tested on the other 169

4. Relevant Information:

   Samples arrive periodically as Dr. Wolberg reports his clinical cases.
   The database therefore reflects this chronological grouping of the data.
   This grouping information appears immediately below, having been removed
   from the data itself:

     Group 1: 367 instances (January 1989)
     Group 2:  70 instances (October 1989)
     Group 3:  31 instances (February 1990)
     Group 4:  17 instances (April 1990)
     Group 5:  48 instances (August 1990)
     Group 6:  49 instances (Updated January 1991)
     Group 7:  31 instances (June 1991)
     Group 8:  86 instances (November 1991)
     Total:   699 points (as of the donated datbase on 15 July 1992)

   Note that the results summarized above in Past Usage refer to a dataset
   of size 369, while Group 1 has only 367 instances.  This is because it
   originally contained 369 instances; 2 were removed.  The following
   statements summarizes changes to the original Group 1's set of data:

   #####  Group 1 : 367 points: 200B 167M (January 1989)
   #####  Revised Jan 10, 1991: Replaced zero bare nuclei in 1080185 & 1187805
   #####  Revised Nov 22,1991: Removed 765878,4,5,9,7,10,10,10,3,8,1 no record
   #####                  : Removed 484201,2,7,8,8,4,3,10,3,4,1 zero epithelial
   #####                  : Changed 0 to 1 in field 6 of sample 1219406
   #####                  : Changed 0 to 1 in field 8 of following sample:
   #####                  : 1182404,2,3,1,1,1,2,0,1,1,1

5. Number of Instances: 699 (as of 15 July 1992)

6. Number of Attributes: 10 plus the class attribute

7. Attribute Information: (class attribute has been moved to last column)

   #  Attribute                     Domain
   -- -----------------------------------------
   1. Sample code number            id number
   2. Clump Thickness               1 - 10
   3. Uniformity of Cell Size       1 - 10
   4. Uniformity of Cell Shape      1 - 10
   5. Marginal Adhesion             1 - 10
   6. Single Epithelial Cell Size   1 - 10
   7. Bare Nuclei                   1 - 10
   8. Bland Chromatin               1 - 10
   9. Normal Nucleoli               1 - 10
  10. Mitoses                       1 - 10
  11. Class:                        (2 for benign, 4 for malignant)

8. Missing attribute values: 16

   There are 16 instances in Groups 1 to 6 that contain a single missing 
   (i.e., unavailable) attribute value, now denoted by "?".  

9. Class distribution:

   Benign: 458 (65.5%)
   Malignant: 241 (34.5%)


wristsavr — forces you to take a break.

wristsavr saves your wrists: it periodically zwrites & xlocks your screen to remind you to take a 2 minute break. This is a little ditty that I whipped up last night to avoid working on my operating system.

usage: wristsavr [-hb] [-m mins]
    -h        Display usage information.
    -b        Bully mode.  If xlock terminates before two minutes,
              xlock the screen again.
    -m mins   wait mins minutes between wristsavr notices (defualt 45)


Java TreeGraphics

A java package for visualizing binary trees in ASCII text. Only the abstract superclass TreeGraphics, and concrete subclasses NullTreeGraphics and ASCIITreeGraphics have been ported from the Pascal source developed for Brown University Computer Science 16.

The TreeGraphics routines work by getting from you a pre-order traversal of your tree, in terms of calls to DrawInternal and DrawLeaf. The exact semantics of these calls are

      DrawInternal(String nodeLabel);

For example:

                                |                                                                         |
                                |                                                                         |
                                25                                                                        45
              .-----------------+-----------------.                                     .-----------------+-----------------.
              |                                   |                                     |                                   |
              |                                   |                                     |                                   |
              19                                  35                                   [_]                                  51
     .--------+--------.                 .--------+--------.                                                       .--------+--------.
     |                 |                 |                 |                                                       |                 |
     |                 |                 |                 |                                                       |                 |
     13               [_]               [_]                38                                                     [_]               [_]
 .---+---.                                             .---+---.
 |       |                                             |       |
 |       |                                             |       |
[_]     [_]                                           [_]     [_]

Remember that TreeGraphics needs the calls in pre-order (root-left-right), so the sequence of calls to create this tree would have been the following:




An untraceable, universally verifiable voting scheme

Seminar in Cryptology
Professor Philip Klein
December 12, 1995


Recent electronic voting schemes have shown the ability to protect the privacy of voters and prevent the possibility of a voter from being coerced to reveal his vote. These schemes protect the voter’s identity from the vote, but do not do so unconditionally. In this paper we apply a technique called blinded signatures to a voter’s ballot so that it is impossible for anyone to trace the ballot back to the voter. We achieve the desired properties of privacy, universal verifiability, convenience and untraceability at the expense of receipt-freeness.

Full text: voting.pdf (Adobe Acrobat PDF, 47K)


xmsg uses Tk/Tcl and Sun RPC to pop up windows of text to a remote user. It is loosely based on the old cs project xmesg, which required you to munge with your xhost. xmsg instead uses a client-server paradigm to avoid security holes.

Unfortunately, before I could finish xmsg, the cs dept. discovered zephyr (a project at MIT), which is much better than xmsg could ever be. Thus, I never finished the project.

source (gzip’d tarfile).

Jacksonville Charter School

Jacksonville Charter School was my group final project for ED/0100: Going to High School in America, 1945-Present.

“What would an ideal school look like if it were built around thematic units that cross disciplines? This paper spells one answer out in detail” (SDSU’s Interdisciplinary Teaching with Technology).

Jacksonville Charter School
Education 100
Profs. Theodore Sizer and Paula Evans
TA Jessica Manlin
November 21, 1995

Special thanks to:

Reference Librarian Michael Jackson
without whom none of this would be possible

Also to:

Jessica, our devoted teaching assistant
T-Bone, our homeboy and inspiration
Evans, G-Bone, Perkins, Hirsch, Dewey, Cushman
the beautiful women of New Dorm 307A
Rachel and Gabe
the exquisite women of 530 Grad Center B
the mysterious upper floors of the CIT
Rock A9
the 3rd floor New Dorm lounge
Andy, who unknowingly paid for all of our photocopies
the androgenous powerbook Reggie
all of our moms and dads
Phoebe and the crickets, may they rest in peace

And everything that kept us going through it all:

the Pool
Viking Onslaught
Every member of every band that we listened to
Dunkin Donuts Munchkins
the one-armed man (we were scared!)
and the allure of “Plan A”

Iuvens non est meus filius
– The Jackson Five


Halloween stories about the Los Altos School District

Los Altos, California is a small San Francisco Bay Area town that made national news back in October 1995 when the school district decided to ban Halloween celebrations. The following stories appeared in the San Jose Mercury News.

The most pertinent of these are the 10/17/1995 front page article and the 10/18/1995 local article. They show that the school board actually reached a really moderate decision.