There is a company in my city which does supply chain optimization, data mining and predictive modelling. I tried hard but was not able to enter there.

Now I want to do similar sort of work. They apply many optimization techniques which makes the decision making of company easier.

Presently I know Python, Django, PHP, Linux and web development.

Can anyone guide me that what i need to do for that sort of work, what thing I need to learn so that after 5-6 years I can also start making those type of softwares for small shops. I mean which languages and tools I need to learn.

  • A small tip: If you want to get into data mining, you might want to learn SQL (very important if your data mining data source is an SQL database), and possibly some SQL-based data mining tools. Oracle has some good ones. A non-SQL tool I used that I remember being decent was Weka (cs.waikato.ac.nz/ml/weka ) Jul 22 '11 at 14:19

As Freiheit said, the language isn't that important. You can work on those problems in Python quite happily.

You should read up on combinatorial optimisation problems. Start with wikipedia very high description of many problems. Typical problems you'll want to look at traveling salesman, bin packing and knapsack problems.

Highly recomend the book: Parallel Metaheuristics for Combinatorial Optimization (1999)

It's a bit old, but it'll open up the domain to you.

  • Is there any tutorial where someone has explained how to implement those algos in programming langauage rather than only math
    – user824981
    Jul 22 '11 at 14:27
  • The book recomended actually deals with the computing side of things quite well rather than focusing on pure maths.
    – Jonno
    Jul 22 '11 at 15:13
  • is that a complete book or its pdf article of 21 pages
    – user824981
    Jul 22 '11 at 16:33
  • It's a full book, couple hundred pages
    – Jonno
    Jul 23 '11 at 22:02
  • Are u sure?? because i could not find it on amazon.com. did u have that book??
    – user824981
    Jul 24 '11 at 17:27

It may not be as much about the languages and tools as it is about the ability to understand and solve the problems.

Of the software engineers I know, the ones that are the most valued (and the best paid) are the ones that can solve problems. When a boss or customer comes to them and says, "I need to do X with my data Y and we're having these problems and challenges" those developers are the ones that can then come up with a solution to that problem in general software terms. They can then code that solution in Java, Rails, PHP, C, .NET, or any other modern language. It is akin to a structural engineer being able to build a bridge out of wood, iron, steel, stone, or aluminum. Its his design that matters more than his materials.

The point is that you may know programming languages, but can you readily apply those languages to solving problems?

  • Agreed. It's more about the math of the problem domain than it is about programming. Jul 22 '11 at 16:47

Not the answer you're looking for? Browse other questions tagged or ask your own question.