Mutation Probability Research Articles

Optimization Model and Algorithm of Cigarette Distribution Route Based on Cluster Analysis

Cigarette distribution is an important part of the tobacco logistics supply chain, and it will affect the distribution cost, efficiency, and service quality. In this paper，we choose a two-stage optimization method to analyze the cigarette distribution route across the Administrative Region in China. First, we use a K-means clustering algorithm to generate an initial center. Then, we optimize the clustering region with the vehicle load and workload as constraints. Finally, we establish the cigarette distribution route optimization model by taking the lowest distribution cost as the objective function. In the model calculation, we not only select the adaptive genetic algorithm as the solution algorithm but also use the scanning algorithm to generate the initial population ,design adaptive crossover and mutation probability, and also design inverse operator to correct the error, and we wish to improve the solution performance of the algorithm by this way.In addition, taking the cigarette distribution of Chongqing Tobacco Logistics as an example, this paper analyzes the number of vehicles, average loading rate, and total distribution distance and distribution cost, and finds that the cross-regional joint distribution route is better than independent distribution. And we also find that using clustering analysis and the genetic algorithm could reduce the amount of computation and solve the problem effectively.

Effect Evaluation of Electronic Health PDCA Nursing in Treatment of Childhood Asthma with Artificial Intelligence.

Asthma in children has a long duration and is prone to recurring attacks. Children will feel chest tightness, shortness of breath, cough, and difficulty breathing when they are onset, which has a serious impact on their health. Clinical nursing is of great significance in the treatment of childhood asthma. At present, the electronic health PDCA nursing model is widely used in clinical nursing as a common and effective nursing method. Therefore, it is very important to evaluate the efficacy of the PDCA nursing model in the treatment of childhood asthma. With the development of artificial intelligence, artificial intelligence can be used to evaluate the effect of the PDCA nursing model in the treatment of childhood asthma. The BP network can effectively perform data training and discrimination, but its training efficiency is low, and it is easily affected by initial weights and thresholds. Aiming at this defect, this work uses the genetic simulated annealing (GSA) algorithm to improve it. In view of the problems that the genetic algorithm falls into local minimum and simulated annealing algorithm has a slow convergence speed, the improved genetic simulated annealing algorithm is used to optimize the BP neural network, and an improved genetic simulated annealing BP network (IGSA-BP) is proposed. The algorithm not only reduces the problem that the BP network has an influence on initial weight and threshold on the algorithm but also improves the population diversity and avoids falling into local optimum by improving the crossover and mutation probability formula and improving Metropolis criterion. The proposed method has more efficient performance.

Team-Teaching-Based Course Scheduling Using Genetic Algorithm

Scheduling problems occur in various fields, e.g., education, health institutions, transportation, sports, etc. Main scheduling problems in education is course scheduling which creates schedules for students and lecturers. In this study, course scheduling allocates the lecturers in the form of team teaching and courses into the class and a certain time to even out the workload of lecturers per day and a group of students per day in one week without breaking the constraint. The method used in this research is a genetic algorithm where Universitas Bhayangkara Jakarta Raya as the case study. The genetic algorithm process is done by getting several candidate solutions that undergo a process of selection, mutation, and crossing over to produce chromosomes with the best fitness values. The objective function in this research is minimizing the average variance of the workload of lecturers and students per day in one week. The parameters used in genetic algorithm are determined based on the Design of Experiments mechanism (DOE). The optimal parameter values ​​used to run the program are as: population size = 50, with probability of crossing over = 0.4 and probability of mutation = 0.008. The results of scheduling with genetic algorithms show that the value of the workload variance lecturers and students by considering team teaching is better than actual scheduling. The application of the genetic algorithm method results in a decrease in the standard value deviation of the workload of lecturers and a group of students in one week is 0.114 (3.68%) and 3.11 (55.7%). In addition, course scheduling uses a genetic algorithm with consider team teaching better than genetic algorithm without considering team teaching because there is no class schedule that clashes in real conditions.

Applying AI Tools for Modeling, Predicting and Managing the White Wine Fermentation Process

This paper reveals two of the challenges faced by Romania and proposes a sustainable and simple solution for its wine industry. First, substantial areas with vineyards that may produce qualitative wine, and second, the very low digitalization rate of industrial sectors. More precisely, this work proposes a solution for digitalizing the fermentation process of white wine, allowing it to be adapted for other control techniques (i.e., knowledge-based systems, intelligent control). Our method consists of implementing a pre-trained multi-layer perceptron neural network, using genetic algorithms capable of predicting the concentration of alcohol and the amount of substrate at a certain point in time that starts from the initial configuration of the fermentation process. The purpose of predicting these process features is to obtain information about status variables so that the process can be automatically driven. The main advantage of our application is to help experts reduce the time needed for making the relevant measurements and to increase the lifecycles of sensors in bioreactors. After comprehensive simulations using experimental data obtained from previous fermentation processes, we concluded that a configuration that is close to the optimal one, for which the prediction accuracy is high, is a neural network (NN) having an input layer with neurons for temperature, time, initial substrate concentration, and the biomass concentration, a hidden layer with 10 neurons, and an output layer with 2 neurons representing the alcohol and substrate concentration, respectively. The best results were obtained with a pre-trained NN, using a genetic algorithm (GA) with a population of 50 individuals for 20 generations, a crossover probability of 0.9, and a probability of mutation of 0.5 that uniformly decreases depending on the generations, based on a beta coefficient of 0.3 and an elitist selection method. In the case of a data set with a larger number of variables, which also contains data regarding pH and CO2, the prediction accuracy is even higher, leading to the conclusion that a larger data set positively influences the performance of the neural network. Furthermore, methods based on artificial intelligence applications like neural networks, along with various heuristic optimization methods such as genetic algorithms, are essential if hardware sensors cannot be used, or if direct measurements cannot be made.

Research on the Application of Artificial Intelligence Wireless Network Technology in the Optimization of University Resources

With the increase of teaching resources in universities, allocation for teaching resources has gradually attracted people’s attention. How to optimize and reorganize the numerous teaching resources so as to maximize the teaching efficiency and achieve the best teaching effect is a question that the majority of college management workers must think deeply. Due to the development of computer science, artificial neural networks and wireless networks have made considerable progress. This article optimizes the resources of colleges and universities from these two aspects. First, this work proposes an improved artificial neural network to optimize school resources. The algorithm pointed out that the adaptive genetic algorithm itself has problems such as slow evolution speed and poor individual diversity of the population. After that an improved adaptive genetic algorithm is proposed to optimize BP network training process. By introducing individual evolutionary trend adjustment parameters, a new calculation formula for crossover probability as well as mutation probability is constructed. Improved BP algorithm can effectively avoid the shortcomings of local minima. Second, additional momentum and adaptive learning rate strategies are used to optimize BP network to promote the efficiency for network training. Third, the artificial intelligence network is deployed in the wireless network, and the wireless network will complete the work of data collection, processing, and transmission, so as to optimize the resources of colleges and universities.

The Detection and Verification of Two Heterogeneous Subgroups and a Risk Model Based on Ferroptosis-Related Genes in Hepatocellular Carcinoma.

# Background. Because of the heterogeneity of hepatocellular carcinoma (HCC) and the complex nature of the tumor microenvironment (TME), the long-term efficacy of therapy continues to be a clinical challenge. It is necessary to classify and refine the appropriate treatment intervention decision-making in this kind of tumor. Methods. We used “ConsensusClusterPlus” to establish a stable molecular classification based on the ferroptosis-related genes (FRGs) expression obtained from FerrDb. The clinical features, immune infiltration, DNA damage, and genomic changes of different subclasses were evaluated. The least absolute shrinkage and selection operator regression (LASSO) method and univariate Cox regression were utilized to construct the ferroptosis-related prognosis risk score (FPRS) model, and the association between the FPRS model and HCC molecular characteristics, immune features, and immunotherapy was studied. Results. We identified two ferroptosis subclasses, C1 with poor prognosis and a higher proportion of patients in the middle and late stages infected with HBV and HCV, having higher DNA damage including aneuploidy, HRD, fraction altered, and the number of segments, and higher probability of gene mutation and copy number mutation. FPRS model was constructed on the basis of differentially expressed genes (DEGs) between C1 and C2, which showed a higher area under the curve (AUC) in predicting overall survival rate in the training set and independent verification cohort and could reflect the clinical characteristics and response to immunotherapy of different patients, being an independent prognostic factor of HCC. Conclusion. Here, we revealed two novel molecular subgroups based on FRGs and develop an FPRS model consisting of six genes that can help predict prognosis and select patients suitable for immunotherapy.

CRISPR Detection and Research on Screening Mutant Gene TTN of Moyamoya Disease Family Based on Whole Exome Sequencing.

Moyamoya disease (MMD) has a high incidence in Asian populations and demonstrates some degree of familial clustering. Whole-exome sequencing (WES) is useful in establishing key related genes in familial genetic diseases but is time-consuming and costly. Therefore, exploring a new method will be more effective for the diagnosis of MMD. We identified familial cohorts showing MMD susceptibility and performed WES on 5 affected individuals to identify susceptibility loci, which identified point mutation sites in the titin (TTN) gene (rs771533925, rs559712998 and rs72677250). Moreover, TTN mutations were not found in a cohort of 50 sporadic MMD cases. We also analyzed mutation frequencies and used bioinformatic predictions to reveal mutation harmfulness, functions and probabilities of disease correlation, the results showed that rs771533925 and rs72677250 were likely harmful mutations with GO analyses indicating the involvement of TTN in a variety of biological processes related to MMD etiology. CRISPR-Cas12a assays designed to detect TTN mutations provided results consistent with WES analysis, which was further confirmed by Sanger sequencing. This study recognized TTN as a new familial gene marker for moyamoya disease and moreover, demonstrated that CRISPR-Cas12a has the advantages of rapid detection, low cost and simple operation, and has broad prospects in the practical application of rapid detection of MMD mutation sites.

Stochastic modeling and availability optimization of condenser used in steam turbine power plants using GA and PSO

AbstractThe prominent objective of the present study is to optimize the availability of condenser units in steam turbine power plants. For this purpose, a novel stochastic model is proposed by considering a condenser as a system consisting of seven subsystems. All the time‐dependent random variables associated with failure rates followed exponential distribution while repair rates are arbitrarily distributed. Furthermore, to attain the optimum value of system availability nature‐inspired algorithm has been applied. Predicting the future phenomena is very difficult and most of the existing algorithms have some limitations like stuck at local minima, and slow convergence rate, and so forth. So, in this study proposed model has been optimized using a genetic algorithm (GA) and particle swarm optimization (PSO)The expressions for steady‐state availability of the system have been assessed using Chapman‐Kolmogorov differential‐difference equations developed by applying the Markov birth‐death process. The supplementary variable technique has been adopted to solve differential‐difference equationsFor a comparative analysis, the steady‐state values, GA, and PSO results are compared. It is revealed that the GA gives the maximum value of availability as 0.9886 corresponding to crossover probability 0.65, population size 50, evaluation 300, and mutation probability 0.5. While using PSO optimum availability value is 0.998152 at 27 iterations, population 55, inertia weight 1, damping ratio 0.85, personal best 1.5, and global best 2. So, PSO results outperform the steady‐state and GA results. It is revealed that mixed bed filter and ejector are the most sensitive components of the condenser those influences most the system effectiveness.

An integrated security hardening optimization for dynamic networks using security and availability modeling with multi-objective algorithm

The dynamicity of today’s networks has created uncertainties for security administrators about the defense options to deploy. In this paper, we consider the administrator’s challenge of selecting and deploying the best set of heterogeneous security hardening solutions for dynamic networks given multiple constraints (such as fixed budget, availability of hardening measures, performance degradation, non-patchable vulnerabilities, etc). The current state of the art does not focus on the dynamic characteristic of modern networks, where the effectiveness of defenses is affected by changes in the networks. Hence, we approach this challenge by developing an integrated method to effectively compute optimal defense solutions for dynamic networks given multiple objectives and constraints. The proposed approach works in the following five phases: (1) input/data collection, (2) model construction (using temporal-graph-based security model and Generalized Stochastic Petri Nets), (3) defense evaluator (based on security metrics), (4) Pareto optimal set evaluator (using Non dominated Sorting Genetic Algorithm), and (5) optimal solution evaluator (based on Weighted Sum Model). To demonstrate the feasibility of the proposed approach, we use a real-world case study while taking into account both the vulnerabilities that are patchable and non-patchable. We investigated the sensitivity of the model parameters based on the dynamic network, and the result showed a good result for values from 0.1–0.4 for the mutation probability while the crossover did not change for all the values. Furthermore, we compared the dynamic network optimization results to a static network and the evaluation shows that our proposed approach could aid a security administrator in selecting the best defense options to deploy for modern networks that are dynamic, with at least 75% and 62.50% of the defense Pareto points appearing in consecutive and in all network states, respectively. Moreover, it also provides an insight into the benefit of each defense option before the deployment.

Diversity behavior in a community model with spatial heterogeneity

Habitat heterogeneity plays an essential role in the processes of generating diversity. Another critical factor is the resource distribution in a community. Here we investigate the diversity patterns by introducing a computational model with spatial structure. Spatial heterogeneity is introduced through the distribution of resources, which is made using a fractal landscape constructed using fractional Brownian motion. In this way, we can adjust the landscape roughness by varying the Hurst exponent. In our model, a species is characterized by a set of half-saturation constants, so each species uses each resource differently. We investigate the evolution of the species number over time. We observed that the species-area relationship has two power-law regimes. The diversity relation with the Hurst exponent depends on the mutation rate and distribution of half-saturation constants. Species diversity does not depend on spatial heterogeneity for high mutation rate. However, a positive relationship is verified for low mutation probability.

Enhanced Black Widow Algorithm for Numerical Functions Optimization

Black widow optimization algorithm is a recently evolutionary metaheuristic that imitates the unique mating behaviour of the black widow spiders in the real life. The trend of published papers utilizing the BWO algorithm is growing rapidly due to its efficiency in solving various engineering optimization problems. However, the BWO does not always perform as well as it should, and this is due to the random initialization of the spiders also the loss of good candidate solutions during the search. To remedy these problems, we propose in this paper a modified black widow optimization algorithm (MBWO) based on three ideas. First, an efficient initialization technique is adopted, which can guarantee starting the search with finest quality spiders and plays a significant role in determining an optimal or near-optimal solution. Second, the sexual cannibalism phase is modified to avoid the loss of high-quality solutions. Finally, an adaptive adjustment of crossover and mutation probabilities is presented to achieve a compromise between the diversification and intensification. Experiments are carried out on nineteen standard benchmark functions with different dimensions. The simulation results reveal that MBWO algorithm outperforms the original one also other metaheuristic algorithms in term of solution accuracy, global optimality, and the convergence speed.

Kingman’s model with random mutation probabilities: convergence and condensation I

AbstractFor a one-locus haploid infinite population with discrete generations, the celebrated model of Kingman describes the evolution of fitness distributions under the competition of selection and mutation, with a constant mutation probability. This paper generalises Kingman’s model by using independent and identically distributed random mutation probabilities, to reflect the influence of a random environment. The weak convergence of fitness distributions to the globally stable equilibrium is proved. Condensation occurs when almost surely a positive proportion of the population travels to and condenses at the largest fitness value. Condensation may occur when selection is favoured over mutation. A criterion for the occurrence of condensation is given.

Improved binary gaining–sharing knowledge-based algorithm with mutation for fault section location in distribution networks

AbstractFault section location (FSL) plays a critical role in shortening blackout time and restoring power supply for distribution networks. This paper converts the FSL task into a binary optimization problem using the feeder terminal unit (FTU) information. The discrepancy between the reported overcurrent alarms and the expected overcurrent states of the FTUs is adopted as the objective function. It is a typical 0–1 combinatorial optimization problem with many local optima. An improved binary gaining–sharing knowledge-based algorithm (IBGSK) with mutation is proposed to effectively solve this challenging binary optimization problem. Since the original GSK cannot be applied in binary search space directly, and it is easy to get stuck in local optima, IBGSK encodes the individuals as binary vectors instead of real vectors. Moreover, an improved junior gaining and sharing phase and an improved senior gaining and sharing phase are designed to update individuals directly in binary search space. Furthermore, a binary mutation operator is presented and integrated into IBGSK to enhance its global search ability. The proposed algorithm is applied to two test systems, i.e. the IEEE 33-bus distribution network and the USA PG&E 69-bus distribution network. Simulation results indicate that IBGSK outperforms the other 12 advanced algorithms and the original GSK in solution quality, robustness, convergence speed, and statistics. It equilibrates the global search ability and the local search ability effectively. It can diagnose different fault scenarios with 100% and 99% success rates for these two test systems, respectively. Besides, the effect of mutation probability on IBGSK is also investigated, and the result suggests a moderate value. Overall, simulation results demonstrate that IBGSK shows highly promising potential for the FSL problem of distribution networks.

The Landscape of Actionable Genomic Alterations by Next-Generation Sequencing in Tumor Tissue Versus Circulating Tumor DNA in Chinese Patients With Non-Small Cell Lung Cancer.

BackgroundCirculating tumor DNA (ctDNA) sequence analysis shows great potential in the management of non-small cell lung cancer (NSCLC) and the prediction of drug sensitivity or resistance in many cancers. Here, we drew and compared the somatic mutational profile using ctDNA and tumor tissue sequence analysis in lung adenocarcinoma (LUAD) and squamous cell carcinoma (LUSC), and assess its potential clinical value.MethodsIn this study, 221 tumor tissues and 174 plasma samples from NSCLC patients were analyzed by hybridization capture-based next-generation sequencing (NGS) panel including 95 cancer-associated genes. Tumor response assessments were applied to 137 patients with advanced-stage (III and IV) NSCLC who first received targeted agents.ResultsTwenty significantly mutated genes were identified such as TP53, EGFR, RB1, KRAS, PIK3CA, CD3EAP, CTNNB1, ERBB2, APC, BRAF, TERT, FBXW7, and HRAS. Among them, TP53 was the most frequently mutated gene and had a higher mutation probability in male (p = 0.00124) and smoking (p < 0.0001) patients. A total of 48.35% (191/395) of NSCLC patients possessed at least one actionable alteration according to the OncoKB database. Although the sensitivity of genomic profiling from ctDNA was lower than that from tumor tissue DNA, the mutational landscape of target genes from ctDNA is similar to that from tumor tissue DNA, which led to 61.22% (30/49) of mutational concordance in NSCLC. Additionally, the mutational concordance between tissue DNA and ctDNA in LUAD differs from that in LUSC, which is 63.83% versus 46.67%, indicating that NSCLC subtypes influence the specificity of mutation detection in plasma-derived ctDNA. Lastly, patients with EGFR and TP53 co-alterations showed similar responses to Gefitinib and Icotinib, and the co-occurring TP53 mutation was most likely to be a poor prognostic factor for patients receiving Gefitinib, indicating that the distributions and types of TP53 mutations may contribute to the efficacy and prognosis of molecular targeted therapy.ConclusionsAs a promising alternative for tumor genomic profiling, ctDNA analysis is more credible in LUAD than in LUSC. Genomic subtyping has strong potential in prognostication and therapeutic decision-making for NSCLC patients, which indicated the necessity for the utility of target NGS in guiding clinical management.

COVID-19 CT image denoising algorithm based on adaptive threshold and optimized weighted median filter

CT image of COVID-19 is disturbed by impulse noise during transmission and acquisition. Aiming at the problem that the early lesions of COVID-19 are not obvious and the density is low, which is easy to confuse with noise. A median filtering algorithm based on adaptive two-stage threshold is proposed to improve the accuracy for noise detection. In the advanced stage of ground-glass lesion, the density is uneven and the boundary is unclear. It has similar gray value to the CT images of suspected COVID-19 cases such as adenovirus pneumonia and mycoplasma pneumonia (reticular shadow and strip shadow). Aiming at the problem that the traditional weighted median filter has low contrast and fuzzy boundary, an adaptive weighted median filter image denoising method based on hybrid genetic algorithm is proposed. The weighted denoising parameters can adaptively change according to the detailed information of lung lobes and ground-glass lesions, and it can adaptively match the cross and mutation probability of genetic combined with the steady-state regional population density, so as to obtain a more accurate COVID-19 denoised image with relatively few iterations. The simulation results show that the improved algorithm under different density of impulse noise is significantly better than other algorithms in peak signal-to-noise ratio (PSNR), image enhancement factor (IEF) and mean absolute error (MSE). While protecting the details of lesions, it enhances the ability of image denoising.

A novel evacuation path planning method based on improved genetic algorithm

In order to solve the problem of finding the best evacuation route quickly and effectively, in the event of an accident, a novel evacuation route planning method is proposed based on Genetic Algorithm and Simulated Annealing algorithm in this paper. On the one hand, the simulated annealing algorithm is introduced and a simulated annealing genetic algorithm is proposed, which can effectively avoid the problem of the search process falling into the local optimal solution. On the other hand, an adaptive genetic operator is designed to achieve the purpose of maintaining population diversity. The adaptive genetic operator includes an adaptive crossover probability operator and an adaptive mutation probability operator. Finally, the path planning simulation verification is carried out for the genetic algorithm and the improved genetic algorithm. The simulation results show that the improved method has greatly improved the path planning distance and time compared with the traditional genetic algorithm.

Context dependency of nucleotide probabilities and variants in human DNA

BackgroundGenomic DNA has been shaped by mutational processes through evolution. The cellular machinery for error correction and repair has left its marks in the nucleotide composition along with structural and functional constraints. Therefore, the probability of observing a base in a certain position in the human genome is highly context-dependent.ResultsHere we develop context-dependent nucleotide models. We first investigate models of nucleotides conditioned on sequence context. We develop a bidirectional Markov model that use an average of the probability from a Markov model applied to both strands of the sequence and thus depends on up to 14 bases to each side of the nucleotide. We show how the genome predictability varies across different types of genomic regions. Surprisingly, this model can predict a base from its context with an average of more than 50% accuracy. For somatic variants we show a tendency towards higher probability for the variant base than for the reference base. Inspired by DNA substitution models, we develop a model of mutability that estimates a mutation matrix (called the alpha matrix) on top of the nucleotide distribution. The alpha matrix can be estimated from a much smaller context than the nucleotide model, but the final model will still depend on the full context of the nucleotide model. With the bidirectional Markov model of order 14 and an alpha matrix dependent on just one base to each side, we obtain a model that compares well with a model of mutability that estimates mutation probabilities directly conditioned on three nucleotides to each side. For somatic variants in particular, our model fits better than the simpler model. Interestingly, the model is not very sensitive to the size of the context for the alpha matrix.ConclusionsOur study found strong context dependencies of nucleotides in the human genome. The best model uses a context of 14 nucleotides to each side. Based on these models, a substitution model was constructed that separates into the context model and a matrix dependent on a small context. The model fit somatic variants particularly well.

Low dose radiation mechanisms: The certainty of uncertainty

This paper reviews the current understanding of low dose radiobiology, and how it has evolved from classical target theory. It highlights the uncertainty around low dose effects, which is due in part to the complexity of “context” surrounding the ultimate expression of biological effects following low dose exposure. The paper makes special reference to low dose non-targeted effects which, are currently ignored in radiation protection and population level risk assessment, because it is unclear what they mean for risk. The view of the authors is that this “lack of clarity” about what the effects mean is precisely the point. It indicates the uncertainty of outcomes after a given exposure. The uncertainty stems from multiple outcome options resulting from the intrinsic uncertainty of the stochastic interaction of low dose radiation with matter. This uncertainty should be embraced rather than eschewed. The impacts of the uncertainties identified in this paper is explored and an approach to quantifying mutation probability in relation to dose is presented.

Probing the mechanisms of two exonuclease domain mutators of DNA polymerase ϵ.

We report the properties of two mutations in the exonuclease domain of the Saccharomyces cerevisiae DNA polymerase ϵ. One, pol2-Y473F, increases the mutation rate by about 20-fold, similar to the catalytically dead pol2-D290A/E290A mutant. The other, pol2-N378K, is a stronger mutator. Both retain the ability to excise a nucleotide from double-stranded DNA, but with impaired activity. pol2-Y473F degrades DNA poorly, while pol2-N378K degrades single-stranded DNA at an elevated rate relative to double-stranded DNA. These data suggest that pol2-Y473F reduces the capacity of the enzyme to perform catalysis in the exonuclease active site, while pol2-N378K impairs partitioning to the exonuclease active site. Relative to wild-type Pol ϵ, both variants decrease the dNTP concentration required to elicit a switch between proofreading and polymerization by more than an order of magnitude. While neither mutation appears to alter the sequence specificity of polymerization, the N378K mutation stimulates polymerase activity, increasing the probability of incorporation and extension of a mismatch. Considered together, these data indicate that impairing the primer strand transfer pathway required for proofreading increases the probability of common mutations by Pol ϵ, elucidating the association of homologous mutations in human DNA polymerase ϵwith cancer.

Customer Segmentation Using K-Means Clustering and the Hybrid Particle Swarm Optimization Algorithm

AbstractIn a competitive market, it is of great significance to divide customer groups to develop customer-centered personalized products. In this paper, we propose a customer segmentation method based on the K-means algorithm and the improved particle swarm optimization (PSO) algorithm. As the PSO algorithm easily falls into local extremum, the improved hybrid particle swarm optimization (IHPSO) algorithm is proposed to improve optimization accuracy. The full factorial design is used to determine the optimal parameter combination; the roulette operator is used to select excellent particles; then, the selected particles are crossed according to their adaptive crossover probabilities; when the population falls into a local optimum, the particles are mutated according to their adaptive mutation probabilities. Aimed at the K-means’ sensitivity to selecting the initial cluster centers, IHPSO is used to optimize the cluster centers (IHPSO-KM). We compare IHPSO with the PSO, LDWPSO, GA, GA-PSO and ALPSO algorithms on nine benchmark functions. We also conduct comparative experiments to compare IHPSO-KM with several conventional and state-of-the-art approaches on five UCI datasets. All results show that the two proposed methods outperform existing models. Finally, IHPSO-KM is applied in customer segmentation. The experimental results also prove the rationality and applicability of IHPSO-KM for customer segmentation.