Archive for the ‘OR Education’ Category

Unified Artificial Intelligence

April 3, 2010

Noah Goodman and others at MIT have recently developed an A.I. based program for determining who would send e-mail to whom at a fictitious company.  However, it’s not the application itself that is of interest.  It’s the way that they carried out the computations.  The program relies on probabilistic rules that get updated over time.  It combines some of the original ideas of expert systems (using rules and implications) with the probabilistic approach that has been successful in recent years.

The research team admits that it doesn’t have a final solution on how this should be accomplished.  It currently is too computationally intensive.

Here are some random thoughts on this issue.

  1. This sounds like an ideal topic for O.R. researchers to get involved in.  We also have expertise in logic and probabilistic reasoning.
  2. Humans are terrible at dealing with probabilities but we are pretty good at forming categories and recognizing patterns.  Perhaps the A.I.  system could be improved by using “flawed human reasoning.”  For example, the system could be overconfident, just as humans are.  It could rely too heavily on recent information.  It could use very simple rules for updating probabilities.  It could rely heavily on other program’s expertise, just as we rely on what other people think.
  3. Perhaps the best way of developing a “unified approach” is to concurrently use several approaches to arrive at conclusions, and refer to each approach as “an expert.”  Then use a unifying approach that takes the expert opinions and arrives at a group opinion.  I think that this is already done, but perhaps in slightly different ways than I am suggesting.

The advantage of making papers harder to understand

September 27, 2009

When I was a college student, a professor of logic once asked us to write a brief paper on a specified topic, which I no longer remember.  I do remember that I didn’t understand the question that we were supposed to address and had no clue what to write.  So I carried out an experiment.  I wrote something such that each sentence appeared to make sense, but so that the paper did not make any sense when taken as a whole.  (Think “Sarah Palin’s speeches” but with crisper logic.)  My experiment was successful as far as I was concerned.  I got a “B.”  My conjecture is that it would have taken the professor too much time to see through my lack of logic, and it was easier to give it a reasonable grade.  (Warning to current students:  do not try this approach with me.)

In today’s Boston Globe, there is a brief article entitled  “You’d sound smarter if you wrote less clearly.” It’s about David Hakes, an economics professor, who simplified a complex argument in mathematical economics.  Hakes explained, ”We managed to reduce the equations in the paper to six. At this stage the paper was perfectly clear and was written at a level so that it could reach a broad audience.”   When the paper was rejected for being “self evident”, Hakes and his co-author decided to make his work less readable and much more complex mathematically, with no added value.  Hakes later wrote ”I personally could no longer understand the paper.”  It was now acceptable to the referees and published.

Hakes’ point extends to Operations Research and other fields that rely on mathematics.    Referees often do not like simplicity because it may make a paper sound “obvious”, and no referee wants to accept a paper that is obvious.  Mathematical complexity usually seems unobvious, and makes the paper more palatable.

As a referee, I have sometimes asked for a substantive revision of a complex paper if I came up with a much simpler version.  For example, I once asked  an 80 page paper to be shortened to around 15 pages, with no loss of content.   But simplifying a paper is, in general, very difficult and is not the task of a referee.  The unfortunate truth is that authors really can increase their odds of publication by making their papers unnecessarily complex.

A revolutionary device for taking notes

July 18, 2009

Recently I heard of a new product called “SmartPen.”  It’s manufactured by LiveScribe.  As far as I can tell, it is a totally revolutionary product for taking notes. You can write notes into a note book and record the audio of what is happening at the same time.  Later, when you click on a word in your written notes, it will replay the audio that was recorded, starting five seconds before the word was written in your notes.  So, if you take notes during a lecture, you can instantly play back any part of the lecture that you want by just clicking on a word.

Here is a detailed blog entry from an MIT undergraduate named Michael Snively.  He writes:  “This pen, hands down, is the most amazing and potentially beneficial note-taking tool I’ve ever seen in my life.”

I haven’t used the product, but plan on buying one for my daughter, who will be starting college this coming fall.

A combinatorial problem on coincidences

May 10, 2009

In response to a recent editorial by Dick Cavett on coincidences,  Jack (4th commenter) responded with the following story:

A relative was married and there was a sit-down dinner for all guests at the reception, including my father, mother, sister and me. I suppose there were 150 people. The hostess wanted totally arbitrary seating, so she placed 150 little cards in a basket, numbered 1 -150, and each guest was to reach in a pick a number as they entered the room. Then she randomly numbered the 150 seats at the tables. It made for some confusion, as each person had to search for his or her seat. In case you haven’t guessed, the four of us where seated in a row. I mean, what are the chances!

You can try solving it yourself before reading on.  Assume that there were 15 circular tables, each with 10 persons each.

Solution. The assignment of people to numbers can be ignored since the probabilities are exactly the same for every possible assignment of people to numbers.  The analysis assumes that the assignment of people to numbers has already been carried out.

There are 150! (that is, 150 factorial) ways of placing the 150 numbers at the tables.  There are 10 ways of selecting four consecutive seats at any specified table, and thus 150 different ways of selecting four consecutive seats at one of the 15 tables.  If the family sits together, this leaves 146 seats for everyone else, and these numbers can be assigned to seats in 146! ways.   Thus, the number of configurations in which the four family members are seated in a row are

146! (150)(4!),

where the 4! is the number of different ways the four family members can be arranged within the four seats.  Therefore, the probability of the four family members being together is

146!(150)(4!)/150! = 24/(149*148*147),

which is approximately 1 in 135,000.  Incidentally, the probability is independent of the number of tables or the size of tables, so long as there are at least 5 persons at each table.   For example, the probability is the same if there was a single circular table with 150 seats.

Please, don’t use Google hit counts to bolster arguments.

April 1, 2009

Don Kleinmutz, the President of INFORMS, wrote an interesting article on the failure of risk management in the recent issue of Analytics Magazine.  But to bolster a clearly justifiable argument, he wrote: “So are these pervasive and systematic failures of risk management something new or unusual?  No.  An Internet search of ‘risk management’ plus ‘failure’ produced 12 million hits on Yahoo! and 4 million hits on Google.”
First, it’s impossible to understand what the number of hits mean.  If the number of hits were 1/10th as large, would this be less compelling evidence?

More to the point, we have no clue what was on the pages or if it supports his conclusions, especially the last 99.99% of the pages. For example, I randomly selected the 5th page of results (which is in the top .002% of all hits), and looked at the first item.  It is entitled “Failure reduction in manufacturing systems through the risk management approach and the development of a reactive maintenance model.”
It’s tempting to use Google for justifications and it’s becoming widespread.  But it is also a bad idea. I can prove it.  I just typed in “Google” AND “bad idea” and got 1.84 million hits.

The beginning of the end for non-interactive, in-class lectures.

March 26, 2009

YouTube has just launched This is a site where educational videos are stored, including lots of videos of lectures.   I see this as the beginning of the end for university lecturing.  Others more prescient than I saw this coming years ago.

By lecturing, I am referring to the mostly one-way communication where a learned faculty member presents a selected topic to a group of students.   The lecture is clearly of value.  But how valuable is it for the faculty member to do it live?  I predict that within a few years, we will learn that it is not very valuable.  It certainly isn’t valuable enough to be the faculty member’s most time-intensive contribution to a subject each semester.

Fortunately, this is potentially a huge win for both faculty and students.  If faculty free up enormous amounts of time by not having to modify old lectures and repeat them in class, then faculty can devote the time in other ways to enhance a subject.

Stretch Goals for INFORMS

March 20, 2009

I encourage the INFORMS leadership to come up with a list of stretch goals.  Each goal should be plausibly achievable, and each goal should be worthy of achievement.  Here is my list.


  • Increase the membership of INFORMS to 15,000.
  • Increase the attendance at the annual practice meeting to 2,500.
  • Actively engage at least 80% of student chapters in activities that benefit the profession as a whole.
  • Ensure that at least 90% of all students enrolled in graduate programs of OR are members of INFORMS.
  • Do what it takes so that the retention rate of members in INFORMS is at least 90% every year.
  • Develop a successful series of well-utilized on-line (and free) tutorials for continuing education for our professionals.
  • Facilitate the creation of executive education programs that are heavily OR-based, especially in-house programs that can more heavily influence the activities of companies.
  • Help develop very wide awareness for OR as a profession that knows how to assess metrics that are difficult to measure and that knows how to solve problems in complex and uncertain environments.
  • Help ensure that at least 80% of American CEOs should know something about OR.
  • At least 80% of all MBA programs should have a very popular subject on “quantitative modeling for qualitative insights” (to paraphrase Steve Powell). INFORMS should facilitate the course creation.
  • All (or almost all) students leaving an MBA program should understand the potential benefits of operations research.

What OR Professionals can do

March 20, 2009

I looked again at the “Science of better” site for INFORMS.  I really don’t like the style, but feel that it is possibly the basis for a much better site.  It’s written in the style of a software company who is trying to get you to buy their software.  It’s not a style that I want for our professional society.  Put another way, The Science of Better site needs to be improved.

Here is a quote from the site.
O.R. is unique. It’s best of breed, employing highly developed methods practiced by specially trained professionals. It’s powerful, using advanced tools and technologies to provide analytical power that no ordinary software or spreadsheet can deliver out of the box. And it’s tailored to you, because an O.R. professional offers you the ability to define your specific challenge in ways that make the most of your data and uncover your most beneficial options.

Here is an alternative that I just wrote.  Admittedly, it is much longer, which makes it difficult to compare the two.   But I hope that it opens a discussion about what is the right tone.  (I am not at all wed to the specifics of the alternative, which undoubtedly could be improved.)

The goal of Operations Research professionals in business is to provide improved decision making in an uncertain and complex environment.   OR professionals draw from a large body of well-developed analytic tools including statistics, simulation, spreadsheet analysis, behavioral decision theory, and optimization.  The OR Professional can benefit your company in a variety of  ways.

  • They can help provide your company a common language (and understanding) for discussing uncertainty and complexity;
  • They can provide analyses of current planning options while helping management to develop new options;
  • They can develop techniques (often simple and inexpensive) for assessing important metrics such as costs, risks, customer satisfaction, and employee morale.
  • They can assess the value of obtaining new information for strategic decisions.  This permits management to seek out new information and data analyses that will have the largest positive impact on their decision making.
  • They have analytic tools for addressing business issues that have a very high level of complexity, uncertainty, and risk.

Let me know if you prefer this type of style, or whether you prefer the style that is currently on the Science of Better site.  And if you have an alternative option, please let me know that.

Continuing Education in Operations Research

March 20, 2009

I just Googled “Continuing Education” AND “Operations Research”.  The top hit was a post on a blog of IE/OR tools that referred back to an earlier post on my blog.  My post was listed 7th on Google.

I’ll admit that I like my posts to be listed high by Google.  But I would much prefer it if it were listed lower because there were lots of other sites about continuing education in O.R.  Continuing education is of vital importance to our profession.  We need to take it seriously.

To be fair, there is a new INFORMS committee that is focused on strategic issues for our profession, including issues relating to continuing education.   I am in contact with the chair of the committee, Cynthia Barnhart.

Is it time to retrain business reporters?

March 16, 2009

As of the time I am writing this, the most e-mailed article from the New York Times is entitled “Is it time to retrain B-Schools“. Kelley Holland began the article by writing “John Thain has one. So do Richard Fuld, Stanley O’Neal and Vikram Pandit. For that matter, so does John Paulson, the hedge fund kingpin.  Yes, all five have fat bank accounts, even now, and all have made their share of headlines. But these current and former giants of finance also are all card-carrying M.B.A.’s.”

Obviously, we now know what caused the financial meltdown.   For your information, their MBAs were all between 29 and 36 years ago.

  1. John Thain, Harvard, 1979.
  2. Richard Fuld, Stern, 1973
  3. Stanley O’Neal, Harvard , 1978
  4. Vikram Pandit, Columbia, 1976
  5. John Paulson, Harvard, 1980.

If only these titans of industry had taken that course in business ethics around 30 years ago, we may have averted this crisis.

In reality, the article has nothing to do with these five infamous MBA graduates except for the initial paragraph quoted above and a quote from the dean of the Thunderbird School of Global Management; rather the article deals with educational issues that have been around for years that business schools are still debating. In fact, the first paragraph was a journalistic version of “bait and switch” so that the reporter could write about general concerns about MBA education.

So, my questions are: Where do New York Times business reporters learn their journalistic ethics? And it is time to retrain them?