Soft Computing - Science topic

Thanks for sharing. Keep it up!

Thanks for sharing.

Hi, Adnan Majeed. The fees are different for each type of participation. Please refer to the official conference website for detailed information: https://ais.cn/u/IVBvuq.

impact factor journal and what is the cost of paper publications ?

Yes, it is feasible to implement user-defined material models in ANSYS and MSC.MARC using C or C++. Here’s a guide on how to proceed with this approach:

ANSYS primarily uses Fortran for user-defined material models (through usermat.f), but you can integrate C or C++ code through the following methods:

MSC.MARC typically uses Fortran for user subroutines (e.g., hypela2.f), but similar to ANSYS, you can integrate C/C++ code as follows:

By using these methods, you can leverage C or C++ for complex material modeling tasks while still using Fortran as the interface within ANSYS or MSC.MARC.

Wishing you every success, International Journal of Complexity in Applied Science and Technology

Dear Mohsen Masoomzadeh ,

Artificial neural networks (ANNs) are machine learning algorithms inspired by the human brain.

ANNs consist of interconnected nodes called neurons, which learn from data to solve problems like image recognition and speech processing. During training, the connections between neurons’ weights are adjusted to minimize errors between the desired and actual output.

Regards,

Shafagat

Dear Kavita Kumari ,

Soft computing techniques have been used for a long time to solve optimization problems. Various disciplines provide real-life problems which are difficult to solve, at least mathematically, in principle; therefore soft.

Regards,

Shafagat

Hi Erkan,

This can sadly happen if the reviewers don't finish their reviews on time. There was already a question asked about the length of the reviewing process in Applied soft computing on ResearchGate:

You can write a polite email to the editor to inquire about the status of your manuscript.

Great, Elaine Lu Happy to Collaborate.

To create an equation based on an ANN, you will need to specify the input variables and the desired output, and then design and train the ANN to learn the relationship between the inputs and the output. This typically involves the following steps:

Once the ANN has been trained and tested, you can use it to make predictions for new data by feeding the input data through the network and using the output from the final layer as the predicted output.

Many companies use a SW static analysis tool, such as Klocwork, Coverity, etc., to get the SW metric. For SW quality, the defect density (how many SW defects found by the code review and/or test (unit test, integration test, etc.)) would be a leading indicator.

Normally, an adaptive controller is designed based on one of the available techniques. Each technique is originally designed for a specific class of dynamic system. The controller is then adjusted as data are collected during run time to extend its effectiveness to control a large class of dynamic systems.

I have reviewed several manuscripts in this journal. This journal is relatively quick to respond. It usually takes 2 months for the first round of review, and an average of 3 weeks each time the paper is revised. Of course, it depends on the reviewers.

Akashah Arshad Fuzzy logic is a method of describing and processing uncertain data. Each fact or assertion, such as 'it will rain tomorrow,' must be either true or false in more conventional propositional logic. Nonetheless, most of the knowledge that individuals use to learn about the world is fraught with ambiguity.

I recommend reading the following papers: https://www.sciencedirect.com/topics/engineering/fuzzy-logic-technique

Hi Safial Islam Ayon , Elsevier provides official word templates for some other journals which you could adapt to the particular needs of ASC, like this one:

But you might also want to get in touch with the editors because as discussed in this threat here on ResearchGate, it looks as if you do not necessarily need to stick to a template and that the formatting will be done for you once your article gets accepted:

Anyways, good luck with your proposal!

I think Convolutional Neural Network CNN and its variants are dominant in this topic.

I would like to invite you to have a look on the following links.

SSO: https://www.mathworks.com/matlabcentral/fileexchange/92150-sperm-swarm-optimization-sso

HSSOGSA: https://www.mathworks.com/matlabcentral/fileexchange/92130-hssogsa?s_tid=srchtitle

MOSFP: https://github.com/sh7adeh1990/MOSFP

Thanks for sharing.

Arash Mazidi PSO is also in various classification problems. I particularly use it for Phishing website datasets.

All classification problems are under unsupervised learning category. Specifically, DNN based algorithms fall into this category where image classification techniques are in use. There are several application areas where these techniques are utilised.

No I couldn't. I used Matlab instead for ANFIS. It was quite robust. You could view the architecture, rules etc quite easily. And you could also customize membership function with ease.

Thank you very much Dear Sevcan for your contribution. ANN is widdely used in recognition and classification and prediction .

As long as we want to model all relative concepts (Velocity, heat, age, evaluation, small, high, success, light, clean etc ..) mathematically and according to expert opinion, and Fuzzy will remain popular. In a world where control of a system (Air conditioning, temperature, velocity, intersection etc ..) is not required, there will be no Fuzzy. So fuzzy will always be. I have at least 10 articles from different areas where I use Fuzzy logic.

for very large datasets, 80/20% to 90/10% should be fine; however, for small dimensional datasets, you might want to use something like 60/40% to 70/30%.

The rules are defined during the training phase so that they can be used during the testing phase, since test samples are considered to be unknown.

Dear Hapuarachchige Don Nelaka Shayamal Priyankara

You can use the IDS in a noncentralized process also.

Regards

Of course not !!! Metaheuristics are here to solve the problem of the time.

Metaheuristics are cleverly done since the exhaustive search with the infinite running time is useless especially for large-scale and real-life optimization problems.

Table one of the following reference gives a detailed answer to this issue:

Apart from the foregoing mentioned discussion, all metaheuristic optimization approaches are alike on average in terms of their performance. The extensive research studies in this field show that an algorithm may be the topmost choice for some norms of problems, but at the same, it may become to be the inferior selection for other types of problems. On the other hand, since most real-world optimization problems have different needs and requirements that vary from industry to industry, there is no universal algorithm or approach that can be applied to every circumstance, and, therefore, it becomes a challenge to pick up the right algorithm that sufficiently suits these essentials.

Well, it is not that efficient solution but you can try to help the model by limiting the number of states, for instance, eliminiting use cases that you think are possible to take into consideration by the model but won't help you to reach your final objective. Or by limiting the action space, like the use of the grid world instead of continuous action space in case of navigation...

As per my information the soft computing algorithm provides better results than standard algorithm if the feasible space is nonconvex. But my questions about the case where in objective functions and constraints are linear in nature. Moreover, if problem is non linear in nature but the feasible space is convex then what be the best algorithms.

First decide your objective function and their parameters (or dependencies). After that you can apply any nature inspired algorithm.

Bonjour Amira,

Il faut savoir que pour des calculs numériques, on se donne à chaque pas de temps une erreur maximale à ne pas dépasser. Quand on reste dans des cas très simples (poutre en élasticité), tu peux avoir une erreure nulle. Si ton problème augmente en complexité, cette erreur ne sera plus nulle et tu auras plusieurs itérations dans un même pas de temps pour converger vers une valeur approchée. Si tu veux savoir quel logiciel est le meilleur, je te dirai que ça dépend de différents critères: ton expérience avec le logiciel, du type de méthode numérique que tu veux faire (EF, EFE, ED...), les lois de comportement déjà implantées. La possibilité d'importer tes propres lois de comportement si besoin est un plus.

Nothing called best. It depends on the problem domain, quality of data, generalization expected etc. If you suggest a problem, a better algo may be suggested, but there is nothing called the best algorithm.

refer the paper:

It depends on the type of problem

Best Regards H. Naderpour

Please see the following links

Best Regards H. Naderpour

Thank you very much Raael Sir, I am doing Same and trying to minimize Error

It depends of your case. Look articles and book of Anil K. Jain

Machine learning and AI can be considered almost same. The basic purpose of AI is to provide intelligence to machines so that they can behave like human beings. One of the aspects of intelligence is learning. So, Machine learning is a topic under artificial intelligence.

AI may not be dealing with imprecision always.Where as soft computing has no special standing unless uncertainty is involved in the process.

This means the increasing of input neurons in the system

You have a good CV. I recomend to see in the job section of researchgate. There, it is frequently offered posodc positions.

Far too many than it should be !

I recommend to:

1) take a look to this position paper which discusses (in a reader-friendly manner) the source of the issue in our field:

2) move on to serious research topics. As a community, we should orient our effort on a (statistical) analysis and unification of optimization methods under simple technical names, rather than wasting critical time (and harming the credibility of the heuristic optimization area) by inventing new names and creating micro-societies dedicated to cats, chaotic birds, frogs and predating water drops.

Before using ANN, it need a learning which is a process that produces an output that is as close as possible to the desired output by adjusting network parameters , there are two learning methods: Supervised and unsupervised.

Learning rules of ANN includes weights modifiable depending on the input it receives, its output value, and the associated teacher response.

The danger in stating which AI method will be the leader in next years is to think that every problem is a nail and we just have a hammer. This has been the problem with the current emphasis on deep learning.

Each problem is unique and must be solved with the appropriate approach, methodology and tools that are adequate for the problem (just take as example the no free lunch theorem in ML). AI is not ML and includes a multiplicity of algorithms and heuristics that in some instances cannot be compared side by side.

With respect to the two criteria "Time" and "Precision", optimization problems can be classified into two main categories: "Unimodal" and "Multimodal". In a unimodal optimization problem (e.g. sphere modal), only a single optimum (maximum or minimum) solution exists around many other non-optima solutions. For such problems, it is reasonable to check the "velocity" (i.e. "line-oriented search" or "path-oriented search") ability of the algorithm to find such single optimum solution as quickly as possible. Thus, here, "Time" is critical. For a multimodal optimization problem, however, many "local optima" exist, trapping the "global optimum single solution". For such problems, it is more reasonable to check the "volume-oriented search" ability of the algorithm to "precisely" escape form those local traps and to locate "reliably" the near optimal solutions or, in the best case, to find the global optimum solution. It is worth to note that, almost all real-world optimization problems fall into the category of multimodal problems.

The answers to the question for this thread are excellent. There is a bit more to add.

Before starting either a M.Sc. or Ph.D. thesis, it is very important to read published theses by others. Here are examples:

Source of M.Sc. and Ph.D. theses:

Another source of theses:

Read this article!

https://kisslibrary.com/book/5202FDEACE2020569814?utm_source=o35-rdl2_ra-1206_7-2&utm_medium=latestpunjabnews.com&utm_campaign=newtraf&search=Recent+Advances+on+Soft+Computing+and+Data+Mining%3A+The+Second+International+Conference+on+Soft+Computing+and+Data+Mining+%28SCDM-2016%29%2C+Bandung%2C+Indonesia%2C+...+in+Intelligent+Systems+and+Computing%29&x=6200117

Softcomputing comes under the umbrella of AI.

Dear H. Naderpour

The Quantum Artificial Intelligence Lab (also called the Quantum AI Lab or QuAIL) is a joint initiative of NASA, Universities Space Research Association, and Google (specifically, Google Research) whose goal is to pioneer research on how quantum computing might help with machine learning and other difficult computer.

Regards, Shafagat

The following file may be helpful. Best of luck.

Hello, I suggest IJIETAP, Scopus indexed.

Dear H. Naderpour,

you speak of a field of the most vast here in short some titles, links and files in attached.

-Soft Computing in Artificial Intelligence | Young Im Cho | Springer

- 7th International Conference on Soft Computing, Artificial Intelligence ...

- Applications of AI and soft computing for challenging problems in the ...

https://www.sciencedirect.com/science/article/.../S09204105050000...

- IJSCAI 2018 : International Journal on Soft Computing, Artificial ...

www.wikicfp.com/cfp/servlet/event.showcfp?eventid=70058...

- Soft Computing and Its Applications - World Scientific

- Applications of Soft Computing - Recent Trends | Ashutosh Tiwari ...

- Application of soft computing techniques in coastal study–A review

https://www.sciencedirect.com/science/article/.../S24680133163001...

- Soft computing-based decision support tools for spatial data

- Soft computing applications: A perspective view - IEEE Conference ...

ieeexplore.ieee.org/document/8321190/

Best regards

Optimization and influence of pyrolysis factors on bio-oil yield and exergetic effiency of the pyrolysis reactor using the afore- mentioned soft computing tools and

Nonlinear system modeling has attracted great attention in engineering sciences as well in many other fields such as meteorology, economics, biology, etc. This is simply due to two main reasons. First, the presence of nonlinear phenomena in most physical systems is rather the rule than the exception. Second, nonlinear system analysis and control can only be achieved if the mathematical model of the system is known . This model can be constructed using the physical modeling approach which leads to the white box models. This approach consists in decomposing the considered system in different parts (electrical part, mechanical part, heat conduction part, etc.) and then applying the appropriate first principles that govern each part. This approach can clearly be very complex since it requires detailed specialist knowledge in many different domains of science. In addition, the first principle models are often non linear and non stationary and consequently the obtained models are too complex for most practical purposes. System identification represents an interesting solution of constructing models, known as black box models, since it allows to overcome most of these problems. Furthermore, it can be applied to all physical systems. This approach consists in constructing mathematical models from observed input and output data based on five main steps : data collection, model class selection, model structure estimation , model. parameter estimation and model validation. The model class selection represents the corner stone toward the objective of nonlinear system identification because the model accuracy depends on the used model class. Numerous classes have been proposed in the literature for nonlinear black-box models such as Volterra models, NARMAX (Nonlinear AutoRegressive Moving Average model with eXogenous inputs models), nonlinear state space, block-oriented models structures, etc. All of these models are able to approximate any nonlinear system with arbitrary accuracy. The choice of the best class depends on the model purposes : emulation, prediction, simulation, controller design, fault detection, etc. For control purpose for example, a mathematical model must ensure a compromise between accuracy, model extraction cost and parametric simplicity of model in order to ensure the implementation of the controller.

The performance depends on the data sets being used. Please test with the classifiers first, and then interpret.

As our work is limited to pyrolsis modeling and simulations , yet ANN, RSM , are evolutionary process optimization well estabilised , much used to reduce the cost of process developments . i found recent one listed here via google serach

Pyrolysis products from industrial waste biomass based on a neural ...

Traduzir esta páginade YF Sun - ‎2016 - ‎Citado por 6 - ‎Artigos relacionadossee more details on the products of biomass pyrolysis were investigated. A three-layer artificial neural network (ANN) model was developed and trained to simulate and predict the selectivity and yield of gas products. Good agreement was achieved between the experimental and simulated results. The major gas products of ...

Kamasani,

Can you please explain by what do you mean "better"? Do you mean faster computation time, better prediction or categorization?

At any rate, I think you might need to review Support Vector Machine as a machine learning technique.

You can use Guzman et al., (2015) as a reference for comparison

Mohamed

For the new comer or beginners, refer to the basic book of soft computing. For example, Soft Computing by D.K.Pratihar. You may also follow (a little advanced step):

1. Burke, E. K., & Kendall, G. (2005). Search methodologies (pp. 1-17). Springer Science+ Business Media, Incorporated.

2. Deb, K. (2014). Multi-objective optimization. In Search methodologies (pp. 403-449). Springer, Boston, MA.

3. Marler, R. T., & Arora, J. S. (2004). Survey of multi-objective optimization methods for engineering. Structural and multidisciplinary optimization, 26(6), 369-395.

Hi Akinjide,

To learn about ANFIS, you should have some knowledge in Neural Nets and Fuzzy Logic. Since ANFIS is invented by Prof. Roger Jang, I'd his article and book “Neuro-Fuzzy and Soft Computing-A Computational Approach to Learning and Machine Intelligence” on this subject:

If we make or think any research (as in some funded projects) as close ended, latter it may be opened anytime.

Hi Lohita,

First important point is what Mr. Frederic Lepretre. You have to know well what you mean about DNA structure and what you want to analyze. Then, choose the method(s).

Thanks

To give some examples of big data analytics under the new paradigm:

Any comments are more than welcome.

if you use these terms in your paper, maybe there is no difference for reviewers 

For example, in MATLAB, scripts are stored in m-files. You can edit your base ANN script in order to include GRNN functionality. 

Dear Paul,

Thank you for the response. I'll take it in consideration.

Dear Kavita Sahu,

You may use ensemble methods of soft computing techniques. There are different way to ensemble  the soft computing techniques are available in literature. You may also use some heuristic techniques in soft computing techniques for improving the performance of the model.

Please follow this papers:

If you need some help, please contact through below mention emil id:

Dear Mahdieh,

The applications of soft computing approach have proved two main advantages:(1) it made solving nonlinear problems, in which mathematical models are not available, possible and (2) it introduced the human knowledge such as cognition, recognition, understanding, learning, and others into the fields of computing. This resulted in the possibility of constructing intelligent systems such as autonomous self-tuning systems, and automated designed systems. This paper highlights various areas of soft computing techniques.

In summary, the advantages of employing soft computing is its capability to tolerate

imprecision, uncertainty, and partial truth to achieve tractability and robustness on simulating human decision-making behavior with low cost. In other words, soft computing provides an opportunity to represent ambiguity in human thinking with the uncertainty in real life

For more on this topic, please read the article entitled "ON SOFT COMPUTING TECHNIQUES IN VARIOUS AREAS " by Santosh Kumar Das et al. Rupak Bhattacharyya et al. (Eds) : ACER 2013, pp. 59–68, 2013. © CS & IT-CSCP 2013

and  SOFT COMPUTING TECHNIQUES (Chapter-3) Zedah L. A. (1992)

Hoping this will be helpful,

Rafik

Please have a look at GeNIe (http://www.bayesfusion.com/, free for academic teaching and research use).  It allows for creating Bayesian networks with continuous nodes/distributions with any relationships between them, not restricted to linear relationships.

Buse, R. P., & Weimer, W. R. (2010). Learning a metric for code readability. Software Engineering, IEEE Transactions on, 36(4), 546-558.

Posnett, D., Hindle, A., & Devanbu, P. (2011, May). A simpler model of software readability. In Proceedings of the 8th working conference on mining software repositories (pp. 73-82). ACM.

Your problem statement is not clear. I do not understand what the (x,y) pair is supposed to represent (a node in the network?) and I do not understand what you mean by "fitness" of the node. Fit for what, exactly? What "resource" are we trying to find? What makes one (x,y) location better than another one?

A genetic algorithm is not good for finding a needle in a haystack; it needs some sort of relatively smooth and mostly connected landscape, such that it can start at a random place and then feel its way in a random fashion to take steps uphill to a peak (or in an inverted landscape downhill to a lowest point).

You need to describe your landscape before you start devising your algorithm. Figuring out how to make your landscape suitable for a GA will tell you what your fitness algorithm is: In fact, every GA should begin there, with a landscape that gives you a fitness metric as a single floating point number, and then you can figure out the specifics of the mating and generational dynamics.

Your data size is very small. Try to collect some more data and they do analysis.

Good morning Aarti 

Yesterday we were running your data (in my AI class) and here we have the results:

As you may know, the anfis structure generated by matlab, uses clustering techniques or the initial generation of membership functions in the antecedent, and fixing these parameters uses a least squares methodology to obtain the consequent parameters. Nevertheless you should be aware of the nature of data by running first the clustering algorithms in order to determine a “good” set of parameters for training. I will lead you through several recommended steps:

 

1. Normalize the data without using the genfis/anfis commands, just normalize and take the data to the hypercube [0, 1] or  [-1, 1]

2. Select randomly a 60% of the data for training, 20% for test and 20% for validation VERY IMPORTANT. 

3. Run clustering algorithms over the whole set via subclust command in order to know the sensitivity of your data. 

You have just 524 patterns but 34 inputs and at least 3 outputs and some clustering algorithms such as subtractive provides 340 clusters with radii 0.5 or 31 clusters with radii 0.2 which means that your data is too sensitive to clustering algorithms. We like to run it several times looking for a nice parameter... but if you know too much the set you can semi-supervise the results.

4. Run fuzzy c means algorithm with the optimal number of clusters from subtractive and try to plot the 3-d projection of the centroids and clusters to verify if the results have any sense for you. I was just wondering that you work on power systems.

5. When you are satisfied with the number of clusters, use just the training set and run the genfis2 command and remember that you are going to build a MISO sugeno structure. Use the options menu to set a nice parameter for subtractive clustering!!!!

6. Also, try a genfis3 and remember to use the fuzzy c means right number of clusters as parameter and also try building a ‘mandami’ type of FIS (see the command line setup) because we find out a nice results with it. Also build a ’surgeon’ structure in order to compare.

7. Now we have 3 fis that reflects the training set with good accuracy, use evalfis command to evaluate antis results and see if it is ok with your research.

8. Try evaluating the test and validation set in order to check if the ANFIS can extrapolate the data

9. it is very important to realize that this structure is highly dependent of the number of patterns to properly converge to the right connections in the network. Try to use not too much clusters because you will be overtraining the net and will never reach the capability for extrapolation.

10. Finally you can train a 'sugeno’ type structure with anfis command but… it seems that genfis2 or  genfis3 can provide you a nice results. 

I will send the code to help you to compare your findings, but I really prefer that you try to replicate this exercise.

 

My kindest regards,

 

Lucia

If I understand the question correctly then you want to know whether to calculate the IGD value in objective space or decision space. Then the answer is in objective space. So the IGD values are calculated between your obtained Pareto front and the reference Pareto front. It can be confusing because Pareto front and Pareto set are sometimes used as synonyms but sometimes Pareto set means optimal decision variables and Pareto front optimal objective values. Hope this helps.

Regards,

Miha

Hello;

Genetic Algorithm is an optimization technique with which you can optimize objective functions or parameters

in your case, Genetic Algorithm can help you to optimize parameters that you are looking the best that they can. I think it's a good optimizer and it work off-line.

This is really an insightful answer. I appreciate your efforts. Now, I will try to incorporate your suggestions and try to see the output in each scenario.

Hi Yves,

Maybe this paper where ANN is compared among other models (MARS, PPR, SVR, RF) is useful to you.

Cheers,

Manuel