Simulated Annealing and the Genetic Algorithm - Coursework Example

Global Optimisation Techniques Coursework 1 Date of production Supervisor: Student name Student number University of ---------- Table of Contents The initial proposal 3 The objective and purpose of the Study 3 Proposed Research Methodology 4 Summary: 6 Literature Review: 7 Genetic algorithms 7 Simulated annealing 8 Review of a Key Paper: 9 Evaluation and Outcomes: 10 Recommendation for Further Investigation 11 Legal, Social, Ethical, Professional Issues and Academic Misconduct: 12 Poster 15 Abstract 15 Methodology: 15 Personal Recommendation for Further Investigation: 16 Conclusions 16 The initial proposal The project Global Optimisation Techniques aims to investigate and explore application areas for Simulated Annealing and the Genetic Algorithm. These techniques are used in for finding the optimum value for an objective function. . While simulated annealing is a random algorithm method genetic algorithm uses regression. Genetic algorithm requires genes to be represented in a manner that they will fit a function such as regression function where genetic operators help in obtaining optimal solutions. This two methods share an important characteristics with other optimisation techniques in that they are primarily search techniques, but unlike the gradient-based search methods, these techniques are largely heuristic and involve large amount of computation and some statistical reasoning. The objective and purpose of the Study The objective of the study is to explore and investigate of the use of Global optimisation techniques that are Simulated Annealing and the Genetic Algorithm. Additionally, this investigative study will also highlight characteristics of the two methods. The various methods will be described as well as compared and contrasted. Objectives The dissertation aims to fulfill the following objectives: • Literature survey of autonomous learning methods that provides an understanding and description of the methods found. • Comparison and contrast of learning methods against each other. • Final recommendations for most appropriate Global optimisation techniques, and areas for further development. As a problem statement, it can be highlighted that the Global optimisation techniques are currently attracting a lot of research interest and new application areas are being explored. Simulated Annealing and the Genetic Algorithm involve large amount computation for analyzing. The data is entered in form of algorithm or probabilistic patterns for the purpose computation. Proposed Research Methodology The methodology would be qualitative methods, which involves literature reviews and document analysis along with a review of the available published material. The primary data cannot be obtained directly from the users; neither can it be obtained from the other authorities since time limit, alternative secondary sources will have to be utilized for the purpose of this research. The methodology can be classified as a mixture of generic qualitative and descriptive design with the different elements of the literature review and the data analysis contributing to the study of the proposed topic. The point here is that the methodology needs to take into account both elements as only then would a suitable basis be arrived at for the secondary source collection and the subsequent analysis from the sources would be according to the design principles that are prescribed for such cases. Finally, the methodology would be validated against established procedures for such studies and care would be taken to stick to the practices of international research studies as much as possible. Data Collection Technique Data will be collected through secondary sources including online libraries, peer reviewed journals, and prior researches closely related to the topic. While the secondary data is general skewed and thus affected the outcome of any research based on that data, the secondary data in question is not skewed. As such, secondary sources of data especially those found on the World Wide Web will be used to develop a comprehensive picture of topic. The inquiry will therefore constitute selection of articles from published and peer reviewed books and journals as well as credible web sources that contain information on the Simulated Annealing and the Genetic Algorithm. Furthermore, all the cited articles and books selected will be up to date having been published from year 1990 onwards. Similarly, websites used in the study will include certified agency websites, news articles from credible media companies as well as materials from authoritative web sources. Limitation of the research to secondary sources will facilitate analysis of numerous articles from diverse sources and authors on the issue of the Simulated Annealing and the Genetic Algorithm. The report Summary: As a problem statement, it can be highlighted that the Global optimisation techniques are currently attracting a lot of research interest and new application areas are being explored. Simulated Annealing and the Genetic Algorithm involve large amount computation for analyzing. The data is entered in form of algorithm or probabilistic patterns for the purpose computation. The main purpose of Global optimisation techniques is to reduce the time of computation for large amount of data and obtain a better solution. Since they use GPU memory, it is possible to process large and complex data like genetic variability of a population of cattle in certain location. Some models may take large amount of time to process some information but Global optimisation techniques computing helps increasing processing time by changing codes. In order for one to process data using these techniques one will need to decompose the data into small units as well as arranging this data in a manner that algorithms will handle it. Global optimisation techniques allow convenient access to a pool of computing resources that configure and provide minimal interaction to data. Global optimisation techniques are revolutionizing computational and allowing them to tap from computing resources that are extremely powerful through the server. The key lesson is that given a problem decomposition decision, programmers will typically have to select from a variety of algorithms. Some of these algorithms achieve different trade- offs while maintaining the same numerical accuracy. Others involve sacrificing some level of accuracy to achieve much more scalable running times. The cutoff strategy is perhaps the most popular of such strategies. Even though we introduced cutoff in the context of electrostatic potential map calculation, it is used in many domains including ray tracing in graphics and collision detection in games (Kirk and Hwu, 2014 p. 91). Computational thinking skills allow an algorithm designer to work around the roadblocks and reach a good solution. Literature Review: Genetic algorithms Genetic algorithm is good in determining optimal solution such as determining the minimum number of test. Algorithm is selected and used for the purpose of matrix multiplication, division, and subtraction. Genetic algorithm relies on the in-built spatial sensitivity of the array to obtain some of the computational information of data. In Genetic algorithm techniques, several reduced data sets are obtained simultaneously by using an array of receiver coils such as a four-channel phased array coil in order to enhance the sampling speed. The most fundamental requirement for successful Genetic algorithm with regard to hardware is an appropriate array of receiver coils. The number of elements of this array of coils depends on the particular application, which ranges from two to eight elements. The coils must also be arranged systematically in order to attain an acceptable signal to noise ratio. The spatial sensitivities of these coils are also crucial and should be reasonably constant throughout the data capture process. In order to make the sensitivities constant, a rigid arrangement such as cage-like arrangement is used (Xianchun, 2014, p.5). Another important requirement for Genetic algorithm concerns the number of receiver channels and the corresponding coil elements. As a rule of thumb, the number of coil element should be equal to the number of receiver channels. When carrying out Genetic algorithm it is also important to accurately establish the coding effects of receiver sensitivities in order to achieve reliable reconstruction (Xianchun, 2014, p.4). In this case the population is encoded as elements vector that match up to expected solution for optimal problem. The initial population of individuals is randomly generated and then evolves over a number of generations, until convergence. A new generation is created by allocating reproduction trials to the individuals according to their fitness, and then randomly selecting pairs of surviving individuals and applying genetic operators to each pair to create their offspring. The algorithm is hybridized by feasibilizing and hillclimbing procedures. At the end of the evolution, an optimal subset of retests would correspond to the maximum fitness of feasible individuals. Simulated annealing Simulated annealing is a method where an optimum solution is obtained upon entry of a cost function it has elements of probability where by there is a finite set that determines the cost function and it is discreet like. Simulated annealing algorithm begins with a random data and all other information is released randomly after an iteration of the algorithm. The Simulated annealing algorithm integrates a feasibilization technique that focused the search in the feasible regions of the solution space. The elements are; finite set, minimum of the function, , collection of positive coefficients , non increasing function and An initial “state. Review of a Key Paper: Walker, J. Methods in Molecular Biology- Genome-Wide Association Studies and Genomic Prediction. London: Springer Science The paper is experimental and does have considered the legal, social, ethical and/or professional issues. The paper has explored theories regarding the improvement of productivity of beef and milk. The authors have further considered genetic variations of cattle or beef cows as critical in determining the amount of milk or beef that can be obtained from crossing breeding them. The paper centers on bovine genome analysis using R program is important by carrying out analysis of the existing data on genomic DNA in Angus, Santa and Brahman. This attempt is aimed at trying to increase Although, it is often been acknowledged that, a large percentage of beef and milk production is a result of a hereditary familial tendency, precise understanding of the basic genetic developments has been missing. Recent studies in molecular biology, nonetheless, have indicated that inherited genes will affect productivity. This paper will inspect analyse Genotype text files finding out the associations of this cattle. . The genetic algorithm were the setupSNP to prepares the data for manipulation as follows MyData=matrix (scan (file="angus.txt", what= integer (),sep="\t"),50000,1000,byrow=T) This helped in reading the file with Bovine Genome for Angus in the angus.txt this data will need to read line by line and this is done using the following R source code myData=readLines(con="angus.txt") This information is turned into matrix by manipulating it using the following R code myData=readChar(con="angus.txt",nchars=file. info ("angus.txt")$size,useBytes=T) Then, this information needs to be saved in the system for analysis and is done using the following code, save (myData,file="angus.bin") After the information has been save it can be worked in the existing file therefore it was loaded to the workspace using the following R code load ("angus.bin") we set up setupSNP’ and start to work by comparing the frequencies of the SNP among the animals under consideration that is Santa, Angus and Brahman Evaluation and Outcomes: Walker, J. Methods in Molecular Biology- Genome-Wide Association Studies and Genomic Prediction. London: Springer Science The authors intended to find out genetic variations (genome-wide analysis) in Brahman, Angus and Santa. In the study Bovine50Kmap was analyzed for Angus, Santa, and Brahman with the intention of providing resources for cross breeding to produce economically viable cattle. Genetic variation analysis showed that there some genotypes that were variable. The genome in Angus appears have high quality beef which will require the cross breeding. Therefore this paper fills a gap in the literature of genetic variability by investigating the Angus, Brahman and Santa by looking SNP. It is the first analysis the connection between these animals. In the study, SNP was used to identify useful phenotypes in determining meat productivity and grading. The authors verified that ARS-BFGL-NGS-119910 dominant and the in Santa while ARS-BFGL-NGS-28540 is dominant is Angus. They concluded that the results are concurrent with the results obtained by other researchers. The small amount of current genetic data on Santa has made the analysis on rapid-cycling varieties highly useful to gaining an understanding of these highly valuable cattle. For each SNP that was provided had its own model with provided P-values for variability. The analysis has provided provide insight on how genes function thus they have to be significantly enriched to improve productivity. The gene functions enriched in nearby genes include the multicellular organism process, regulation of biological quality, and cell morphogenesis. This analysis provides estimated results of expected functions, so additional study of function consequences between actual phenotypes should be carried out. Recommendation for Further Investigation After an in-depth literature review on previous studies and research conducted in this study, it was clear that research on Global Optimisation Techniques remains an exhausted and requires more researches. Nevertheless, the study gained enough information likely to answer the research questions and hypothesis testing. This research topic however was inexhaustible by this study as well as previous researches because the results keep on conflicting. Therefore, this study recommends use of different materials and research designs that could reveal more information on this topic. However, researchers should not research on some areas that already confirmed by previous studies. Therefore, the study recommends review of previous studies to locate areas that require more research. One of the most exciting new developments is genetic algorithms, if correctly developed, will compute data in a timely and effective manner. If it is programmed in an ineffective manner, data collect and response to its environment will not be beneficial to organizations. Legal, Social, Ethical, Professional Issues and Academic Misconduct: Ethical Consideration It is important to note that careful consideration was given to confidentiality in the composition of the research methodology of this paper. The informed consent form will be included in the online survey and available as a separate introductory page to all participants. It will aim at informing the participants of the background of the study, procedures for participation, confidentiality, the voluntary nature of the study, as well as ethical considerations. This study will rely on an online format for consent that states, “Completion of the questionnaires signifies your consent to this study.” This format avoids receiving returned consent forms with participant names and signatures and therefore allows anonymity of respondents. All participants are free to withdraw from the study at anytime during the research process. This research study does not impose any major risk, threat or harm to participants. This is also made clear in the separate introductory page of the study. Researchers that promise confidentiality to their research participants must keep their data confidential. Human research subjects have a right to privacy, which means that information they provide to the researchers must be kept private. The guidelines of promised confidentiality remain a difficult topic, but as part of informed consent, a statement about the extent of confidentiality of records must be delineated. Moreover, the matter of confidentiality is absent in the Belmont Report nor are the professional codes of different organizations consistent in matters of confidentiality regarding human research subjects. Some professional organizations have provisions about absolute confidentiality while others do not have any statements about it. Still others have very constrained clauses about the confidentiality of private information. Unless the person waives it, some professional organizations have absolute confidentiality, and do not permit researchers to divulge life-threatening, illegal, or abusive situations if revealed. Moreover, they would prevent disclosure even under court subpoena whereas other professional organizations such as American Psychological Association recognize limited confidentiality, and the right of the judicial system to acquire confidential information. Still others describe in detail the circumstances under which confidentiality is required, and researches must reveal the limits of confidentiality if it is limited. Failure to disclose the limits of confidentiality to persons means absolute confidentiality is implied. Unless they stated in informed consent that they would provide information with court subpoena, confidentiality is absolute for survey researchers as well. However, their professional code also permits them to waive confidentiality if they uncover life- or health-threatening situations. The question of ethics in research has always plagued scholars and researchers alike to the extent that it raises important issues. Doing research, whether social or scientific, is fraught with problems that scholars and researchers must resolve. They must decide what they are investigating in their research, and how they will go about testing their hypotheses. They must consider if they want human or animal subjects and follow the code of ethical guidelines for them. They must also assess how much money they will allocate in the budget for their research, what protocols to follow, and how long it will take for them to complete it. Statement of Copyright The text written herein is under the sole ownership of the author, who is protected by copyright laws. No phrase, quote, or any form of material from this thesis should be made part of any other publication efforts unless given the specific consent of the author in a written capacity. Moreover, any material quoted, following the express consent of the author, should be referenced with the name of the author, or relevantly acknowledged. Declaration of originality This is to certify that the work is entirely my own and not of any other person, unless explicitly acknowledged. The work has not previously been submitted in any form to the ---------------------- University or to any other institution for assessment for any other purpose. Poster Abstract The main aim of the paper was to investigate and explore simulated annealing algorithms and genetic algorithms which are commonly used in computation. From the results of the study it can be noted that simulated annealing algorithms has different approach to computation while genetic algorithms has different execution time and user parameters. Methodology: For this research study, researchers will adopt quantitative research method which is based on positivism paradigm which is suitable for collecting and analyzing data pertaining to a natural phenomenon. The justification for this approach is the based on the purpose of the study which is to explore and investigate Simulated Annealing and the Genetic Algorithm and then making objective conclusions. Personal Recommendation for Further Investigation: Further research needs to be carried out on the topic especially on the subject of Global Optimisation Techniques characteristics. Global Optimisation Techniques allows convenient access to a pool of computing resources that configure and provide minimal interaction to data. D Global Optimisation Techniques is revolutionizing computational and allowing them to tap from computing resources that are extremely powerful through the server. The key lesson is that given a problem decomposition decision, programmers will typically have to select from a variety of algorithms. Some of these algorithms achieve different trade- offs while maintaining the same numerical accuracy. Others involve sacrificing some level of accuracy to achieve much more scalable running times. The cutoff strategy is perhaps the most popular of such strategies. Even though we introduced cutoff in the context of electrostatic potential map calculation, it is used in many domains including ray tracing in graphics and collision detection in games. Computational thinking skills allow an algorithm designer to work around the roadblocks and reach a good solution. Conclusions The choice among the five algorithms is dependent on the regression tester's requirements. For example, to test all affected definition-use pairs, despite spending more regression testing time, the algorithms to choose would be slicing and adapted firewall. To choose a minimum number of test cases and, hence, to perform fast regression testing, the selection would be the genetic or simulated annealing algorithms, although they themselves are slow. The incremental algorithm would be the best choice for selecting a number of test cases whose outputs may be affected, while being the fastest among the five algorithms. However, an overall assessment, based on the ten criteria and the examples used in this work, tends to indicate that the incremental algorithm has more favorable properties than the other four algorithms. References Baradhi, G., & Mansour, N., 1997. “A Comparative Study of Five Regression Testing Algorithms” Binkley, D., 1992. Using semantic differencing to reduce the cost of regression testing. Proc. Conference on Software Maintenance. Chen, Y., Rosenblum, D.S., & Vo, K. 1994. TestTube: A system for selective regression testing. Proc. Int. ConJ on Software Engineering, 1994,211-220. Gregg, C., 2009. Genetic Algorithms in Autonomous Embedded Systems. Harrold, M. & Soffa, M., 1996. An incremental approach to unit testing during maintenance. Proc. Conference on Software Maintenance. Huber, W., von Heydebreck, A., Sültmann, H., Poustka, A. & Vingron, M., 2002. Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 2002;18 Suppl 1:S96-104. Irizarry RA & Gautier L., 2003. The Analysis of Gene Expression Data: Methods and Software. New York: Springer. Kumar, A., Tiwari, S. & Mishra, K., 2010. “Generation of Efficient Test Data using Path Selection Strategy with Elitist GA in Regression Testing”, Leung, H.K.N & White, L., 1992. A firewall concept for both control-flow and data-flow in regression integration testing. Proc. Conference on software Maintenance,. Lin Chen Ziyuan Wang Lei Xu, 2010. “Generation of Efficient Test Data using Path Selection Strategy with Elitist GA in Regression Testing”, Mansour, N. & Fakih, K. 1997. Natural optimization algorithms for optimal regression testing. COMPSAC'97. Nachiyappan, S., Vimaladevi, A., & SelvaLakshmi, C., 2010., “An Evolutionary Algorithm for Regression Test Suite Reduction”. Pharr,M., 2005. GPU Gems2: Programming techniques for high performance graphics and general-purpose computation. Reading, MA: Addison-Wesley. Rajasekaran, S & Pai, G., 2004. Neural Networks, Fuzzy Logic and Genetic Algorithms: Synthesis and Applications. PHI Learning PVt. Rothermel, G. & Harrold, M., 1993. A safe, efficient algorithm for regression test selection. Proc.Conference on Software Maintenance. Satish, N. Harris, M. & Garland, M., 2009. Designing efficient sorting algorithms for proceedings of the 23rd ieee international parallel and distributed processing symposium. Walker, J., 2008. Methods in Molecular Biology- Genome-Wide Association Studies and Genomic Prediction. London: Springer Science. You, L. & Y. Lu, 2012, “A Genetic Algorithm for the Time AwareRegression Testing Reduction Problem”. Read More

Finally, the methodology would be validated against established procedures for such studies and care would be taken to stick to the practices of international research studies as much as possible. Data Collection Technique Data will be collected through secondary sources including online libraries, peer reviewed journals, and prior researches closely related to the topic. While the secondary data is general skewed and thus affected the outcome of any research based on that data, the secondary data in question is not skewed.

As such, secondary sources of data especially those found on the World Wide Web will be used to develop a comprehensive picture of topic. The inquiry will therefore constitute selection of articles from published and peer reviewed books and journals as well as credible web sources that contain information on the Simulated Annealing and the Genetic Algorithm. Furthermore, all the cited articles and books selected will be up to date having been published from year 1990 onwards. Similarly, websites used in the study will include certified agency websites, news articles from credible media companies as well as materials from authoritative web sources.

Limitation of the research to secondary sources will facilitate analysis of numerous articles from diverse sources and authors on the issue of the Simulated Annealing and the Genetic Algorithm. The report Summary: As a problem statement, it can be highlighted that the Global optimisation techniques are currently attracting a lot of research interest and new application areas are being explored. Simulated Annealing and the Genetic Algorithm involve large amount computation for analyzing. The data is entered in form of algorithm or probabilistic patterns for the purpose computation.

The main purpose of Global optimisation techniques is to reduce the time of computation for large amount of data and obtain a better solution. Since they use GPU memory, it is possible to process large and complex data like genetic variability of a population of cattle in certain location. Some models may take large amount of time to process some information but Global optimisation techniques computing helps increasing processing time by changing codes. In order for one to process data using these techniques one will need to decompose the data into small units as well as arranging this data in a manner that algorithms will handle it.

Global optimisation techniques allow convenient access to a pool of computing resources that configure and provide minimal interaction to data. Global optimisation techniques are revolutionizing computational and allowing them to tap from computing resources that are extremely powerful through the server. The key lesson is that given a problem decomposition decision, programmers will typically have to select from a variety of algorithms. Some of these algorithms achieve different trade- offs while maintaining the same numerical accuracy.

Others involve sacrificing some level of accuracy to achieve much more scalable running times. The cutoff strategy is perhaps the most popular of such strategies. Even though we introduced cutoff in the context of electrostatic potential map calculation, it is used in many domains including ray tracing in graphics and collision detection in games (Kirk and Hwu, 2014 p. 91). Computational thinking skills allow an algorithm designer to work around the roadblocks and reach a good solution. Literature Review: Genetic algorithms Genetic algorithm is good in determining optimal solution such as determining the minimum number of test.

Algorithm is selected and used for the purpose of matrix multiplication, division, and subtraction. Genetic algorithm relies on the in-built spatial sensitivity of the array to obtain some of the computational information of data. In Genetic algorithm techniques, several reduced data sets are obtained simultaneously by using an array of receiver coils such as a four-channel phased array coil in order to enhance the sampling speed.

Read More