There are many research on metaheuristic optimization e.g. Particle Swarm Optimization, Genetic Algorithm, etc. Some study show that they are good for clustering task. But I don't find any comparation about it.
Which one is the best to be applied for optimizing the clustering process?
daer friend
As such there is no best algorithm in data mining.
All things depend on the data.
For a particular data, particular algorithm may be the best.
So you need to do some homework before you embark on any decision.
I hope this helps you.
Best regards
Dr.Indrajit Mandal
Hi Arie,
I totally agree with the above researchers. You should select the algorithm which is best suited to your problem.
Regards,
Rajesh.
Hi Arie,
I acknowledge the all comments from the above. Try some of metaheuristics (SA, GA, PSO, ACO, VNS, etc.) to your problem then find what has the best results.
Hi,
We have a nice book related to the topic:
http://www.springer.com/computer/ai/book/978-3-540-92172-1
Also you can find few papers related to 'adaptive clustering':
http://www.softcomputing.net/publication.html
It is very hard to generalize the performance of any MH algorithm for clustering. So we need to try a few of them and perhaps can even consider ensembles etc.
All the best!
Hi Arie,
We have studied some PSO variants for Community Detection problem in social network. The problem in background actually nothing but the graph clustering i.e. clustering performed on graph data. May be you can find helpful the following paper for your decision.
http://link.springer.com/chapter/10.1007/978-3-319-11218-3_46
It is difficult to say exactly which algorithm is better. Moreover, each of these algorithms have several variants. That is why algorithms selection is very difficult . See the following paper titled
"Swarm Intelligence Techniques and Their Adaptive Nature with Applications"
of following book
http://www.springer.com/engineering/computational+intelligence+and+complexity/book/978-3-319-12882-5
But, yes advanced versions of these algorithm perform really well. Generally, for graph related clustering problems, ACO or combinatorial variants of PSO, GA are considered. For general clustering problems I would recommend you to use advanced variants of PSO, since PSO has very powerful mechanism relevant to clustering i.e. co-operative learning.
I think any optimization algorithm will work well as this problem is not very complex.
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
https://www.researchgate.net/publication/296636431_Codes_in_MATLAB_for_Particle_Swarm_Optimization
Code Codes in MATLAB for Particle Swarm Optimization
https://www.researchgate.net/publication/280244004_A_comparative_study_of_metaheuristic_optimization_approaches_for_directional_overcurrent_relays_coordination
Article A comparative study of metaheuristic optimization approaches...
Arie Wahyu Wijayanto The NFL theorems state that all metaheuristics averaged over all possible objective functions yield the same average results. The question you should ask is: what is the algorithm that works best for me under the constraints that I have imposed on a specific problem or a class of problems.
For added reading, you may refer to our paper here where we have laid out the swarm clustering problem step by step and shown a few case studies:
Conference Paper Data Clustering using a Hybrid of Fuzzy C-Means and Quantum-...
Hope this helps. SSG
The following current-state-of-the-art optimization algorithms give you the answer:
N. K. T. El-Omari, "Sea Lion Optimization Algorithm for Solving the Maximum Flow Problem", International Journal of Computer Science and Network Security (IJCSNS), e-ISSN: 1738-7906, DOI: 10.22937/IJCSNS.2020.20.08.5, 20(8):30-68, 2020.
It has a complete discussion about your question.
Or refer to the same paper at the following link:
https://www.researchgate.net/publication/344324283_Sea_Lion_Optimization_Algorithm_for_Solving_the_Maximum_Flow_Problem
- Badges
- Science topic