Comment

Comment by Vladimir Lukin on March 23, 2010 at 8:26pm

Huayin,

Ha-ha, it is almost a year from your posting, and no threads. You need to package your algorithm and sell it.
I see a lot of examples to apply your solution - to find an optimal offer allocation under quantity constraints. Also, I see how teams of statisticians are trying to solve it forcing SAS/OR, or brute search approaches to come up with "pseudo-optimal" solutions. Spending time and money.
Well, your solution is elegant, fast, efficient and the most important, find the global optimum solution.

Hope to see that your algorithm will be released someday.

Comment by Huayin Wang on March 30, 2009 at 8:40pm

I am a little surprised that no one take up the challenge. Not even a single person from one of the company I mentioned earlier? Come on, stand out and say it if you think your product is the best!

Well, I am going to be the first since no one does. I have developed an algorithm is the absolutely the best, better than any of those I mentioned, algorithm-wise not the UI part of it :)

Do not believe it? I am providing FREE service!! You do not have to tell me what's the best value you get from your objective function; just let me run my algorithm against your data and I will tell you what the max value I get, for FREE.
It is awfully quite here.

Any taker?

Comment by Rafat on March 6, 2009 at 9:30pm

Hi All, it is an interesting subject of discussion. I am in a practical field more than theoretical application on optimization method. To me optimization is a directional indication of the most appropriate solution among many other workable solutions in your feasible region to your problem that maximize (minimize) your objective with meeting all of your constraint. Meanwhile, analytical is a mathematical evaluation of the problem you had.

Comment by Felipe de Jesus Castro Hernandez on March 6, 2009 at 11:26am

Hi, I am Felipe Castro from Mexico, I have been working many years in order to solve optimization problems. To solve some problems you need to know models and programming, if the models are small I suggest you to use MatLab if you want to solve large models and very easy you don´t know how to programming you should use "Risk" you can see it in the web page www.solver.com and you can use it with Excel, it is very easy to use and you can solve large and difficult models, if you need helping please tell me.

Comment by Huayin Wang on March 2, 2009 at 10:43am

I have a question to ask all you in the group, and it is about one particular optimization problem: the marketing offer optimization - or optimal assignment problem as called operation research.
The question is this: of all the current vendors of marketing optimization solutions (from SAS, Austin Logistics, Market Switch of Experian and others), which is the best and what's your experience of it. Feel free to add any vendor that I have not mentioned.
Thanks!