Clustering Using a Combination of Particle Swarm Optimization and K-means

Patel, Garvishkumar K.; Dabhi, Vipul K.; Prajapati, Harshadkumar B.

doi:10.1515/jisys-2015-0099

Cited by 20 publications

(12 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In addition, this paper also presents a comparison of the results produced by the proposed approach and with K-means based on the validity steps of the grouping such as inter and cluster distance, quantization error, silhouette index, and Dunn index. Comparison of results shows that as the size of the dataset increases, the proposed approach results in a significant increase in the K-means partition grouping technique [31].…”

Section: Related Workmentioning

confidence: 99%

Segmentation of Tourist Interest on Tourism Object Categories by Comparing PSO K-Means and DBSCAN Method

Chindyana¹,

Wulandhari²

2021

View full text Add to dashboard Cite

Marketing in travel companies will usually offer promos or recommendations regarding various categories of random tourist objects to their customers. The promo or recommendation contains categories of tourist objects that are frequently visited and had good ratings from many customers. However, because companies do not really know and understand the characteristics or interests of each customer, sometimes some promos do not match their interests so that they are not interested in taking the promos that are offered. There are already several papers that discuss tourism recommendations, but they only focus on 1 tourist spot or tourism object category. Based on these problems, this thesis is made to discuss the segmentation of tourist interest in tourism object categories by comparing the PSO K-Means method and the DBSCAN method, which is about recommendations for more specific tour packages according to rating. Characteristics or similar interests between 1 tourist and other tourists will be grouped into 1 cluster. From each cluster that is formed, it can make it easier for companies to know what categories of tourist objects each customer is interested in or like and be able to offer promos or recommendations for tour packages according to tourist interests.

show abstract

Section: Related Workmentioning

confidence: 99%

Segmentation of Tourist Interest on Tourism Object Categories by Comparing PSO K-Means and DBSCAN Method

Chindyana¹,

Wulandhari²

2021

View full text Add to dashboard Cite

show abstract

“…SVM has high recognition accuracy for single state equipment [21] and is not good at handling multi-state or continuously changing loads. The k-nearest neighbor method is mostly used to solve clustering problems, which is very sensitive to the selection of parameters, and each classification needs to calculate the distance between the unknown data and all training samples, which requires a large amount of calculation [22]. After comparison, the generalized regression neural network (GRNN) has a significant nonlinear mapping ability and strong approximation ability.…”

Section: Introductionmentioning

confidence: 99%

Load Identification for the More Electric Aircraft Distribution System Based on Intelligent Algorithm

2022

View full text Add to dashboard Cite

Accurate identification of electrical load working status can provide information support to the remote electrical distribution system (EDS) of more electric aircraft (MEA), which could use it to realize redundant switching and protection. This paper presents a method to automatically identify the load status on the remote power distribution unit (RPDU) of MEA by using an intelligent algorithm. The experimental platform is built in an aircraft Electrical Power System (EPS) distribution large-scale test cabin. Four pieces of typical aviation equipment are installed in the test cabin and powered from RPDU. Voltage and current values under 15 working combinations on the RPDU are measured to extract the steady-state V-I trajectory. In total, 750 group samples were collected in the feature parameter database. A generalized regression neural network (GRNN) identification model was established, and the smoothing factor was calculated by using a conventional cross-validation method to train and reach an optimal value. However, the identification results are not ideal. In order to improve the accuracy, the parameter of GRNN was optimized by genetic algorithms. The proposed model shows great performance as accuracy of all 15 classifications reached 100%. The proposed model has advantages of flexible network structure, high fault tolerance, and robustness. It can realize global approximation optimization, avoid local optimization, effectively improve GRNN fitting accuracy, improve model generalization ability, and reduce model training calculation.

show abstract

“…Larger values indicate a greater separation between clusters, meaning less overlap between the clusters in the model. The formula for inter-cluster distance is as follows [34]:…”

mentioning

confidence: 99%

“…Smaller values indicate more compact clusters and are therefore desired. The formula for intra-cluster distance is as follows [34]:…”

mentioning

confidence: 99%

See 1 more Smart Citation

Benchmarking Studies Aimed at Clustering and Classification Tasks Using K-Means, Fuzzy C-Means and Evolutionary Neural Networks

Pickens

Sengupta

2021

MAKE

View full text Add to dashboard Cite

Clustering is a widely used unsupervised learning technique across data mining and machine learning applications and finds frequent use in diverse fields ranging from astronomy, medical imaging, search and optimization, geology, geophysics, and sentiment analysis, to name a few. It is therefore important to verify the effectiveness of the clustering algorithm in question and to make reasonably strong arguments for the acceptance of the end results generated by the validity indices that measure the compactness and separability of clusters. This work aims to explore the successes and limitations of two popular clustering mechanisms by comparing their performance over publicly available benchmarking data sets that capture a variety of data point distributions as well as the number of attributes, especially from a computational point of view by incorporating techniques that alleviate some of the issues that plague these algorithms. Sensitivity to initialization conditions and stagnation to local minima are explored. Further, an implementation of a feedforward neural network utilizing a fully connected topology in particle swarm optimization is introduced. This serves to be a guided random search technique for the neural network weight optimization. The algorithms utilized here are studied and compared, from which their applications are explored. The study aims to provide a handy reference for practitioners to both learn about and verify benchmarking results on commonly used real-world data sets from both a supervised and unsupervised point of view before application in more tailored, complex problems.

show abstract

Clustering Using a Combination of Particle Swarm Optimization and K-means

Cited by 20 publications

References 7 publications

Segmentation of Tourist Interest on Tourism Object Categories by Comparing PSO K-Means and DBSCAN Method

Segmentation of Tourist Interest on Tourism Object Categories by Comparing PSO K-Means and DBSCAN Method

Load Identification for the More Electric Aircraft Distribution System Based on Intelligent Algorithm

Benchmarking Studies Aimed at Clustering and Classification Tasks Using K-Means, Fuzzy C-Means and Evolutionary Neural Networks

Contact Info

Product

Resources

About