well, your problem is not so clear. what is the characteristics of the problem? Is that combinatorial or continuous optimization problem? Why do you want to use GA to solve this problem? What is the computational complexity?- is that NP hard or NP complete? Finally there is a no free launch theorem. So even if you are willing to use any other metaheuristics algorithm, you need to design that algorithm to enhance its performance. 

Metaheuristic Algorithms — http://scholar.google.it/scholar?as_ylo=2014&q=metaheuristic+algorithms+scholar&hl=it&as_sdt=0,5&as_vis=1 [but your concrete needs should be explained a bit more than this]. Hope the suggestion could be still hopefully.

—g 

As you know, the scheduling problem can be reduced to the coloring problem which is NP complete. I have used a genetic algorithm (years ago) to solve scheduling problems in manufacturing processes as well as for class scheduling in an entire university: it worked. I used permutations of the task to schedule as chromosomes.You might want to take a look at:

http://www.sciencedirect.com/science/article/pii/S1018363913000135

you can use Evolutionary Algorithms or kind of divide and conquer techniques. It might help.

I would suggest to have a look at Diferential Evolution. This method has less parameters than GA and it has lower computational demand. But it is also a stochastic metaheuristic.

Thank you for your suggestions and sorry for not mentioning problem clearly.

I am trying to schedule classes for my institute ( I have class rooms and multiple batches of students but number of students in each batch is different and strength of each classroom is different. By using GA i want to come up with a solution that will give me best combination of this) 

This is almost same as mentioned by Oscar Chavoya-Aceves.

I will try to implement your suggestions.

If you have any suggestions please continue the discussion.

I converge in having a look to Differential Evolution. As your problem, as said before, is NP-Hard, your best option, to me, is still looking at stochastic metaheuristics.

our team recently worked on a review paper called "Swarm Optimization Algorithms: A Survey". This paper is soon to be submitted to a journal. we compared the fitness performance of 7 evolutionary methods including Genetic Algorithms (GA), Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), Differential Evolution (DE), Artificial Bee Colony (ABC), Glowworm Swarm Optimization (GSO), and Cuckoo Search Algorithm (CSA) on 20 different benchmark functions (Unimodal and Separable, Unimodal and Inseparable, Multimodal and Separable , Multimodal and Inseparable,Multimodal, Unimodal, Separable, and Inseparable). The results indicated PSO as the best performing method followed by DE and CSA. I can't share the paper yet since it is soon to be submitted but I am happy to send it to you as soon as it get accepted.

You should first classify your problem's complexity.  In its most general form it is NP-hard, but that doesn't mean that yours actually is hard.  Take a look at my paper with Michael Carter, "When is the classroom assignment problem hard?", which appeared in Operations Research.  Carter did classroom scheduling for his own university and for Canadian high schools. Take a look at his papers on the subject.  There is a pretty good chance that you can run a linear program and find the true optimum solution in a few seconds of CPU time. 

Maybe you can try some local search based algorithm for the complex problems, such as tabu search, iterated local search, variable neighborhood search, and so on.

If you want to stick to nature inspired algorithms try ant colony optimization, particle swarm optimization or simulated annealing. They all are algorithms for exploring the local solution space. An alternative which requires a bit of math is to solve your problem thru linear programming. It will give you one optimal solution. Why I say one is because multi objective problems have an entire set of solutions lying on the Parero front. You can also go for something simpler such a heuristics or even a backtracking approach.

I managed to transform an advanced scheduling optimization problem into a weighted fair queuing problem, which typically can be solved in polynomial time. The incoming "jobs" to be scheduled are at first placed in several alternative queues, one for each possible set of parameters. When it is time to deliver a new "job", the algorithm chooses the most efficient set of parameters (i.e. the set of queues that gives best performance in next timeslot), and also deletes delivered job from all other queues. In my cellular communication application, I call it PARPS (Packet and resource plan scheduling), but a similar concept may be applied to many other scheduling applications. 

As an example where standard mathematical programming software (Branch & Bound) works very well in practice for a real-life scheduling problem, take a look at this paper:

http://www.optimization-online.org/DB_FILE/2013/08/3995.pdf

As, I feel for solving such types of problems, you may use Particle Swarm optimization and ABC method.

Good alternatives are Differential Evolution, evolutionary strategy, and evolutionary programming. For an appropriate design of GA, the algorithm should provide efficient solution for larger problems than conventional integer programming approaches.

one may alternatively try a beam search algorithm based on a reduced enumeration (solution) tree. by choosing appropriate filters you might be able to discar only "non-prmissing" solutions. we are currently working on such an algorithm  for scheduling job-shop

There are two possibilities:

- using of differential algorithm such as quasi-Newton

- using an evolutionary algorithm such as PSO (particle swarm optimization), Tabu Search (TS), Ant Colony Optimization (AC), etc…

You may use PSO or DE.

You may try mathematical programming approach like integer programming and then choose an appropriate algorithms for solving the integer program.

you may also try some branch-and-bound or for big instances s kind of beam search that will cut all the solutions except the "most promising" say k solutions on every level of the search tree

PSO algorithm is more efficient

MATLAB programming codes for PSO

https://www.researchgate.net/publication/296636431_Codes_in_MATLAB_for_Particle_Swarm_Optimization?ev=prf_pub

https://www.researchgate.net/publication/280244004_A_comparative_study_of_metaheuristic_optimization_approaches_for_directional_overcurrent_relays_coordination?ev=prf_pub

Code Codes in MATLAB for Particle Swarm Optimization

Article A comparative study of metaheuristic optimization approaches...

Alternative is PSO and codes are here

https://www.researchgate.net/publication/296636431_Codes_in_MATLAB_for_Particle_Swarm_Optimization?ev=prf_pub

Code Codes in MATLAB for Particle Swarm Optimization

You may use expert fuzzy system.

Kindly, go through the following publication.

Y. del Valle, G. Venayagamoorthy, S. Mohagheghi, J.-C. Hernandez, R. Harley, Particle swarm optimization: Basic concepts, variants and applications in power systems, IEEE Transactions on Evolutionary Computation, 12 (2) (2008) 171-195.

https://www.researchgate.net/publication/297245624_Particle_Swarm_Optimization_Algorithm_and_its_Codes_in_MATLAB

Research Particle Swarm Optimization: Algorithm and its Codes in MATLAB

I would consider using CP Optimizer (https://www.researchgate.net/publication/265258884_IBM_ILOG_CP_Optimizer_for_Detailed_Scheduling_Illustrated_on_Three_Problems): you focus on a declarative model of your problem (like in Math Programming approaches but with extended mathematical concepts that makes it much easier to capture scheduling constraints) and you let the automatic search solve your problem (the search is an exact algorithm that can prove optimality but it also uses some meta-heuristics to quickly produce good quality solutions).

Data IBM ILOG CP Optimizer for Detailed Scheduling Illustrated on...

Please have a look at the below book "Evolutionary Optimization Algorithms".

http://eu.wiley.com/WileyCDA/WileyTitle/productCd-0470937416.html

Hi Prashant

we published a review in plosone a while ago which compare the performance of most EA methods in addition to their most successful modified versions. this would most likely give you an idea of what is out there and how well they perform.

Article A Comprehensive Review of Swarm Optimization Algorithms

You may try Backtracking Search Optimization Algorithm.

There are tons of algorithms. Here a list:

DE

PSO

ABC

CMA-ES, etc