The rules say: "What kind of questions should I not ask here? No homework questions! And, yes, we can tell."

Are you sure? I don't think you can.

Because actually I do intend to discuss specific worked exercises from my textbooks and online sources. Where else am I going to discuss OR? INFORMS... which is for professionals... OK fine but this here is the one and only INFORMS discussion site, right?

Do you really think I will learn a serious subject without any human contact? I am smart, but really, you flatter me unduly if you do think that. I am not working currently. I am a stay at home Dad for the time being.

Listen up: I study. I am not a student at some school. I do not have an advisor. But, I have books including textbooks (BELIEVE IT). I read books (BELIEVE IT). I work exercises from books (BELIEVE IT). I fully intend to use what I learn on the job, for money, as I have already done in a dozen different subjects (BELIEVE IT).

Don't believe me yet? OK here is one specific example. I studied calculus and linear algebra the summer after I graduated from college, and because of this I passed a Society of Actuaries math exam. This passing score of a hard exam, directly resulted in several job invitations at insurance and consulting companies being sent to me. The way I studied was to solve the questions from previous exams. There was no advisor. Unfortunately, when I got stuck, I got stuck! There was no internet yet. I made do. Today we have the internet, so we can do better.

Don't believe me yet? OK another. I studied relational database theory using a university textbook starting one year after I graduated from college. I was a software developer by occupation by this time. But, I only have a math degree and I did not take any course in relational databases at school. I just went to the Rutgers University bookstore and bought whatever book from the shelf, that they were using for teaching people about relational databases at the university. During the day I was designing a FoxPro database system to be used by a program manager at the US Army at Fort Monmouth, so I had to learn fast. The system and others I went on to develop were a big success, and my boss told me I was the star employee and I did win several large pay raises over the course of a few years doing this. In fact, it turns out that I studied that university textbook on relational database design well enough to have since personally designed relational databases and star schema databases for data warehouses for several major banks and major insurance companies and smaller companies too. I make my money by software development even though I have just a math degree. That's because I taught myself a dozen serious subjects AFTER I graduated from school, and I used textbooks to do it. There was no advisor.

Here is a third example. There was no web programming, no ASP, no ASP.NET, no Java, when I was in college. These things did not exist yet. So of course I taught myself web programming, java programming, visual basic, and C sharp, javascript, html, xml, using .... wait for it.... books. For several years thereafter I was a professional web developer, and a good one, judging by the money I have accumulated at banks and insurance companies and tech companies some really famous ones of which you undoubtedly have heard.

Here is a fourth example: I recently taught myself parallel programming, over the course of three years. Do you think this subject is simple? Well it's not simple. I used the textbook by professor Gregory Andrews, "Multithreaded, Parallel, and Distributed Programming". I read his book cover to cover, SLOWLY (it is not possible to read fast) TWO times. I worked several of the exercises in nearly every chapter. Additionally I taught myself C and Python and OpenMP and MPI along the way besides. I used to be a windows programmer, now I'm a capable unix programmer as well. I fortunately got to know Mr. Andrews as a correspondent. Unfortunately I did not pay the University of Arizona any money, although I certainly would love to have officially taken the course which Greg designed, if I had had the opportunity to formally enroll. But I could not enroll. My family and I live on the east coast, and I am in my 40s, and I am established and have kids and property here, and I had to work at the time. I learned concurrent programming on shared memory and distributed memory machines the harder way, by studying in near-isolation, and programming lots and lots of code, and writing lots and lots of pencil and paper designs in my now-full 5 subject notebook (full of just one subject: parallel programming). I also studied all the MIT and Berkeley online course videos in this particular subject. It's not as good as talking to an instructor or a teaching assistant, but we make-do.

At this point I am hoping everyone might be starting to believe that people can read textbooks without "homework" being assigned to them.... I do "homework" of my own free will. Learning is not an option; it is mandatory way of life for information workers, particularly software developers and OR/MS "students". These fields require hard study and real school is just not possible for many of us from time to time.

People learn, believe it or not, even though they are (way) beyond formal school. It would be a ridiculously narrow world view to insist that only students read books. As for me, I would never have learned a dozen or so different technical subjects after I graduated from college 20 years ago if I did not continue to use books in a very serious manner. Learning is a fact of life in the information age for knowledge professionals, and books are a big and necessary part of learning. Books are often the most well thought out presentations of subjects by real experts. You can't google everything that's worthwhile (yet).

That said, with apologies in advance:

With due respect, I however fully intend to post OR textbook exercises including my best models or solutions here. I explicitly want to solicit some collaboration on exercises. This is not "homework" -- I am not attending someone's classes.

Please, please, don't ban my upcoming posts regarding specific exercises from popular textbooks and online problem sources, and please, please don't ban my user account.

I want a place to discuss OR/MS problems, WITH NO SCHOOL and NO ADVISOR GETTING PAID. I do pay for books. You should see my house, too many books. My wife reads too.

I took a few courses in OR and management science over 20 years ago. AMPL did not exist at that time, and laptops are cheap now, and so now we all can write and SOLVE even the NP-hard MIP models we develop as solutions to textbook problems. Big data analytics is also on the rise, especially with regards to advertising and medicine and many other fields. It's a great time to do OR. For OR textbooks I am mainly using Quantitative Analysis for Management (Render and Stair), and Introduction to OR (Hillier and Lieberman). I have another modeling book on the way, thanks to the recommendations provided by the users on this very site.

My crazy theory is that someone else out there wants to learn something along with me. Not everyone has free time and free money to pay for an advisor at some school. We do what we must in spite. The internet is nice. I have a kid. My wife has asked me not to attend school, so that I can care for my kid as number one priority. I have chosen to learn the hard way: no instructor. Please accept this life choice. One day I may enjoy the luxury of paying for a course and an instructor but not at this time.

Please, let the serious learners here discuss any and all OR exercises using the Internet as a medium. The policy makers of OR-EXCHANGE and INFORMS really should not dismiss the potential value of peer collaboration for learning that the internet, and this site, technically enables -- if the policy of this site would permit it.

I have been learning to use AMPL for OR. I use the GLPK implementation on Linux. SPECIFICALLY, I want to post up my ideas about some exercises in Chapter One of the free MIT online textbook. This is the textbook which was highly voted as being among the best known, and Mr. Trick himself cited it too.

Thanks for reading

Geoffrey A.

asked 19 Feb '13, 17:02

GeoffreyA's gravatar image

accept rate: 0%


To be completely honest: I haven't (yet) read more than the title and the first paragraph of your post.

Nevertheless, – to my best knowledge – the 'usual' practice" over here is: if a question is well-posed, shows some 'involvement', etc.pp. it'll be left open for discussion; if it's just sth. like "Hey, I have this graph with five nodes [actually, it's from texbook XYZ, but I don't tell you], can you compute its max flow for me?", it probably won't.

(19 Feb '13, 17:18) fbahr ♦

If someone is part of the community and shows a real willingness to learn, then this community has shown itself to be incredibly helpful and insightful. But when "CheatingStudent015" with karma 1 posts the 50th question of the form "Here is an LP: solve it for me and I need it by 5AM tomorrow", the community is harmed. The rule is there for the protection of the community, which is the paramount goal of any rule.

I don't think you can point to a single instance where someone was really willing to learn and was turned away here. But the "no homework" rule certainly does let us keep the community limited to those willing to learn, rather than to cheat their way towards a degree.

So, ask your questions: you might get chided if the questions veer too much toward the "give me the answer", but you have to do a lot more than that to get yourself banned.


answered 19 Feb '13, 19:53

Mike%20Trick's gravatar image

Mike Trick ♦♦
accept rate: 17%

I agree with most of the answers here. Like the others said, the "no homework rule" here is really more of "no cheaters" rule. I'm not sure if folks can always "tell", but as have been alluded to, there are usually indications (they are very blatant).

If you leave a explanatory note saying you're an independent learner and demonstrate that you've put in some effort (i.e. listing the attacks you have tried [which also help folks who answer to eliminate red herrings and zero in on the actual answer], etc.) , I'm sure you'll get a good response. This community is one of the most helpful and least judgmental ones that I've seen in the web Q&A space, and as I understand it, this was a conscious decision borne out of the experiences of the journeymen from sci.math.operations-research.

That said, tons of copy-and-pasted textbook questions can decrease the signal-to-noise ratio. When finding a good juicy question becomes too tedious, folks who answer may simply not bother. The whole thing then becomes a lose-lose proposition, and the quality and utility of this website will deteriorate.

So, there has to be some sort of balance, otherwise the community will cease to thrive.


answered 19 Feb '13, 23:19

Gilead's gravatar image

Gilead ♦
accept rate: 17%

edited 19 Feb '13, 23:33

Does there have to be some sort of balance because the OR community is small? Math Overflow allows only research-level questions, but has a large user base. If the community was large enough, would it be better to have a similar system to the math QA sites, with mathoverflow for research questions and math.stackexchange for lower-level questions?

(20 Feb '13, 01:25) Luis de la T...

That's a good observation. I'm not sure if a MathOverflow / math.stackexchange type arrangement would work for a specialized topic like OR. If we were to limit OR-Xchange to research/practical questions, we might only get a few people talking (as it stands, we are already observing the Pareto principle at work here). On the other hand, if we turn into a StackOverflow, a lot of questions will go unanswered because unlike them, we don't have a critical mass of people to answer questions. I don't know what the solution is, except maybe to invent a mechanism to replicate @Paul Rubin. ;-)

(20 Feb '13, 11:19) Gilead ♦

@Gilead: What makes you think I'm not already a mechanism (or, more precisely, a Watson-based spambot)?

(21 Feb '13, 16:39) Paul Rubin ♦

@Paul: Whenever (either) AI is that advanced - or people are so "dumbed down" to not notice - ...send me a memo.

(22 Feb '13, 04:18) fbahr ♦

First, let me congratulate you on your persistence. Believe it or not I think your question is one of the longest ones on OR-X. Getting back to the issue at hand, here is my own thought about this issue as both a member of OR-X and a lecturer in academia. I'm sure others including @Michael can provide you with more better reasons.

I think the key is "And, yes, we can tell" part of the website FAQ. Over the past two years, I have seen multiple instances of someone asking a very specific homework question (in some cases, you could even find the question on the web). Funny thing is that the person asking the question is usually in a hurry too. Those questions would not get an answer as some students are trying to use the OR-X to solve his/her homework with limited or no effort. In fact, no university professor or lecturer want his/her students to just search or ask in forums for answers to homework. However, we also believe that not every body has access to the same high quality resources as well as good professors and TAs. If some body ask a homework-type question but shows that he/she has worked enough to gain some insights towards the solution, we would be more than happy to help. Although, we are supposed to wait for some days before answering the question.

Finally, remember that if you have really worked on your problem, you won't just ask a homework question in whole. You would probably ask a very specific question and that would make us certain that we should help you (of course, if we can).


answered 19 Feb '13, 17:40

Ehsan's gravatar image

Ehsan ♦
accept rate: 14%

edited 19 Feb '13, 17:41

This is a very interesting thread. From reading it (and in particular @Ehsan's answer), I conclude that we ask nothing more than we ask from students coming to our office hours: you can ask questions as many as you like, but they must not be of the general sort "I did not get that with Adam and Eve; could you please start all over?", but rather specific, well-posed questions. Then, you are identified as a sincere learner, and you will get answers and discussion.


answered 21 Feb '13, 01:19

Marco%20Luebbecke's gravatar image

Marco Luebbecke ♦
accept rate: 16%

Your answer
toggle preview

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here



Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text]( "Title")
  • image?![alt text](/path/img.jpg "Title")
  • numbered list: 1. Foo 2. Bar
  • to add a line break simply add two spaces to where you would like the new line to be.
  • basic HTML tags are also supported



Asked: 19 Feb '13, 17:02

Seen: 1,598 times

Last updated: 22 Feb '13, 04:18

OR-Exchange! Your site for questions, answers, and announcements about operations research.