Intelligent System Assignment: Article Review On Revolutionary Algorithms

Question

Task:
You are required to select a paper based on the intelligent systems for analytics and write a detailed research report on intelligent system assignment critiquing on the selected paper.

Answer

Introduction
Identification of the paper for preparing intelligent system assignment
Nowadays in the zone of flexible computing research, people can notice a large force on finding or developing techniques which are connected with nature. The study emphasizes upon an article which is basically about how to find techniques or enhance techniques with the focus on revolutionary perspectives. The whole family of it can be named as evolutionary computation algorithms [9]. In their territory, there are different algorithms. The main ones are known as genetic algorithm (GA), genetic programming (GP), differential evolution (DE), the evolution strategy (ES) and then evolutionary programming (EP). All of them have their diversity and are also used in various industries. This report is about the various types of processes based on intelligent systems for Analytics.

Purpose of the report on intelligent system assignment
The main purpose of the study has been to display a small survey on the active use of these processes. This is an addition to how in real life the swarm intelligence is put to work. The study highlights various types of algorithms or processes along with the research method used and all problems or issues that are faced with the use of these types of techniques.

Discussion
Content of the article
Genetic algorithm is known to be the oldest and popularly used process which is related to nature [1]. In the present scenario, it has explored solutions that copy the natural process that often takes place within the environment and also the Darwinian Theory is preferred. In this process it's seen to have occupants, who are called chromosomes, they constitute a possible way to fix the issues. Thus, the solving depends on the explanation of the intention. It concerns the dependency of the individual on how much they suit the situation and also the worth of the individual and quality in the fitness of the individual as it is considered to be the main factor. Therefore, highly advantageous candidates are preferred for this case. In this process, the particular article has revealed three partner selection, crossover and mutation.

Genetic programming is a new process and it is a specific form of GA. As evidenced by the study utilized herein intelligent system assignment, it deals with specific types of solutions and methods to improve genetic operators. This was propounded by Koza as a venture to find ways for automatic generation for the program codes taking into consideration their proper function [2]. Here the solution factor dealt with trees instead of chromosomes that are when they try to change the form of trees by mutation of the leaf.

Differential evolution algorithms are mainly known for enhancing the continuous lookout for scope. This process was debated by Storm and Price. There were some drawbacks to this but advantages also exist. The main reason being it is easy to use; it has well-organized ways to memorize things and lowers computer problems and effort. It can give more detailed information on how the framework can be modulated and found. The important prospect of DE is to find out the difference between two individuals chosen from the population. Therefore it stops the solving of forcing a community to the developed functions.

The evolution strategy is different in comparison to GA especially in terms of Thor selection format. In GA, the next generation has been formed by choosing fit individuals and in ES a temporary population which is formed which has a different size [3]. Thus, fitness has not been that important here. It has been evidenced that the individuals go through crossover and mutation. From this process, a certain number of individuals are selected for the next generation. ES works on the vector of the numbers on floating-point, while traditional GA runs on the binary vectors [7]. There are two types of ES (1 + 1), and ES (l + k). The first type is known to be the oldest approach where only one individual has been evolved that is x. The initial is generated causally. In this, the crossover process does not occur but in mutation, it creates another individual y through adding a randomly produced number to each gene of the individual x. The other type uses mutation, recombination and selection to a population of individuals which contain solutions to get better results as much as possible.

Evolutionary programming was formed to be used as an instrument to discover the grammar for an unfamiliar language.

Research method
As mentioned in the article, it has dealt with the main focus being the use of intelligent systems for analytics. It includes the effect of evolutionary algorithms in today's world and how they are used. There were different types of processes which were explained in detail before. Each technique has brought forth in their pseudo-code form that can also be used for its easy execution in any of the programming language. It displayed the main belongings of each algorithm. It also shows a better explanation of the early evolutionary methods and how it was used and put to work. Lastly, it focuses on what issues we tend to face with these procedures in real life [4]. The research for EA has restricted handling of its techniques and is still a hot topic and is applied in today's real life. There has been different open research problems discussed as well. Many practical applications have been franchised by various research places like Honda Research Institute Europe. It is believed that in the years to come researchers might focus on the main drawbacks. The evolutionary algorithm works with the operation research management science. It is considered to be a popular realm in research. Many researchers work towards the EA's to improve their methods. It is said that an interesting domain within future research in EAs is also mimetic algorithms [8].

However, a survey was considered in the present article regarding exploitation and exploration within EAs. Apart from this, certain experiments have been carried out such as the practical applications of state-of-the-art along with evolutionary algorithms as well. The experiments have proved the effectiveness based on the proposed method compared to the back propagation and a neural network model as well as the fuzzy expert system. As evidenced by the present article, it was known to investigating the problem of array in linear dipole of synthesis. However, with the experiments of flattop as well as cosecant squared pattern, with the advantages along with the effectiveness for the particular approach has been verified in comparison to the optimisation and the phase of amplitude that is joint optimisation.

What are the issues or problems identified in the article explored in the segments of intelligent system assignment?
EA's are considered to be an interesting area of research. But they have open problems of research as well like controlling the balance between the exploring and misuse of properties, the self-adaptive control of steering parameters, and also introducing new schemes of selection and expanding their effectiveness [5]. The later fact is that it is important and needed for evolvable hardware. They need more well-organized techniques to control and handle the process. Then they cannot tend to implement the procedure in a noisy environment and need a calm and quiet one to carry out the process. The expenses are also required to be seen because no one can estimate how much can be needed. Also, sometimes to carry out certain sceneries, it needs to check how adaptable the people are along with the state of complexity and difficulty of implementation that tends to be an issue.

Results
As for the future trends in EA's when they cannot use a committed algorithm for a problem that has occurred, it can use one of the EA's which tends to be useful. But when it is done it is important to remember the issues the person might face. The quality of the objective that is chosen tends to have a great influence and effect on the result. Therefore, the main target should be the duplicability as in the access to use the techniques for more than one time and to see that the results also came to be the same with each try. This tends to be the most important factor in all industries [6]. The results are also collected to keep things in check. These results denote the suggested approach tends to quickly satisfy the need. It has also been useful in investigating and scheduling.

According to the results, it has shown that the proposed approach has been quickly able to produce a satisfactory Pareto solution compared with PSO and GA algorithms. The results have demonstrated the planned method which is helpful within the stage of preparation to install flexible large areas for the solar areas.

Conclusion
The report on intelligent system assignment has covered the area of how the intelligence system of analytics affects the lives of people. It has been believed that in the future, new evolutionary algorithms can be created and developed. With time, the problems can also increase and this might seem to be a hot topic then as well. Hence, it is the duty to stay up to date with the changes and be aware of what is happening.

References
[1]H. Yao, "Cloud Task Scheduling Algorithm based on Improved Genetic Algorithm", International Journal of Performability Engineering, 2017. Available: 10.23940/ijpe.17.07.p9.10701076.

[2]M. O’Neill, "Semantic methods in genetic programming", Genetic Programming and Evolvable Machines, vol. 17, no. 1, pp. 3-4, 2015. Available: 10.1007/s10710-015-9254-4.

[3]D. Dubey and A. Mehra, "Pareto-optimal solutions for multi-objective flexible linear programming", Journal of Intelligent & Fuzzy Systems, vol. 30, no. 1, pp. 535-546, 2015. Available: 10.3233/ifs-151778.

[4]A. Sobey and P. Grudniewski, "Re-inspiring the genetic algorithm with multi-level selection theory: multi-level selection genetic algorithm", Bioinspiration & Biomimetics, vol. 13, no. 5, p. 056007, 2018. Available: 10.1088/1748-3190/aad2e8.

[5]O. Grošek and V. Hromada, "Rotation-Equivalence Classes of Binary Vectors", Tatra Mountains Mathematical Publications, vol. 67, no. 1, pp. 93-98, 2016. Available: 10.1515/tmmp-2016-0033.

[6]C. XU and G. PENG, "Fast algorithm for 2D Otsu thresholding algorithm", Journal of Computer Applications, vol. 32, no. 5, pp. 1258-1260, 2013. Intelligent system assignment Available: 10.3724/sp.j.1087.2012.01258.

[7]"Analysis of F5 Algorithm and Improvised F5 Algorithm in Image Hiding with quality safeguarding and Video Hiding by Improvised F5 Algorithm", International Journal of Modern Trends in Engineering & Research, vol. 4, no. 10, pp. 114-128, 2017. Available: 10.21884/ijmter.2017.4317.vxirh.

[8]W. Banzhaf, "Genetic Programming and Emergence", Genetic Programming and Evolvable Machines, vol. 15, no. 1, pp. 63-73, 2013. Available: 10.1007/s10710-013-9196-7.

[9]A. Slowik and H. Kwasnicka, “Evolutionary algorithms and their applications to engineering problems,” Neural Computing and Applications, vol. 32, no. 16, pp. 12363–12379, Mar. 2020.