The BAM (Biarri Applied Mathematics) 2012
The second Biarri Applied Mathematics conference, or BAM, was hosted by the University of Melbourne over two days, November 12 to 13. The conference was a big success, with around 80 people attending, including industry representatives, academics and students.
All the talks (given by Operations Research practioners from both Biarri and from industry, as well as academics) were informative and interesting and generated discussion and questions.
Biarri once again thanks those who attended as well as those behind the scenes who helped make the event a success.
Hi,
I really enjoyed the conference. Here is my 2 cents, written from a student perspective.
I think the BAM conference did, and can continue, to help bridge the gap between theory and practice for OR students. The academic environment is heavily weighted towards the theory. For example in lectures and textbooks the theory of relaxations, decompositions, heuristics etc is only ever accompanied by small “toy” examples. It rarely mentions how these work in practice and never discuss how they are best implemented, which is particularly unfortunate when the ‘devil is in the details’. The BAM conference is in prime position to help fill this void. On this note, the pdf of the bender’s decomposition talk is not included in the downloadable zip, any plans to publish it?
The BAM conference is also well positioned to expand on the difference between producing code and producing software. As students, we write code which will probably never be seen by anyone but ourselves and we can get away with poor programming practices. Specifics on best practices for implementing OR software would be great!
Future topics which I think would be valuable to students include
* The transition from academic to business environment. What are the experiences of your newer staff? Were there any difficulties or surprises?
* How is the process of designing a solution structured in the business environment? Is there a particular methodology that Biarri follows?
* What technologies and software are used to interface with MIP solvers and implement the final solutions? Why are they chosen over their alternatives? Do they differ from those used in the prototyping or proof of concept stage?
* How are MIPs best coded to allow for readability and for possible future extensions? Does the strategy change depending on the depth of control needed over the MIP solver?
Cheers, and thanks again!
You look great guys! You really inspire us.
I love the participation, hope that will improve your business. Continue to post an informative blogs.
Benchmarking Process Improvement