An Improved K-means Algorithm based on Mapreduce and Grid

Ma, Li; Gu, Lingjia; Li, Bo; Ma, Yue; Wang, Jin

doi:10.14257/ijgdc.2015.8.1.18

Cited by 18 publications

(10 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Namun, k-means memiliki masalah dalam sensitivitas penentuan partisi awal jumlah klister. Dapat diketahui bahwa penentuan jumlah klaster sangat penting dalam algoritma k-means (Ma, Gu, Li, Ma, & Wang, 2015). Beberapa artikel (Barakbah & Kiyoki, 2009;Ma et al, 2015;Yadav & Sharma, 2012) menyatakan algoritma k-means sangat tergantung pada penentuan titik pusat klaster awalnya.…”

Section: Pendahuluanunclassified

“…Dapat diketahui bahwa penentuan jumlah klaster sangat penting dalam algoritma k-means (Ma, Gu, Li, Ma, & Wang, 2015). Beberapa artikel (Barakbah & Kiyoki, 2009;Ma et al, 2015;Yadav & Sharma, 2012) menyatakan algoritma k-means sangat tergantung pada penentuan titik pusat klaster awalnya. Untuk setiap percobaan, dengan pemilihan pusat klaster awal secara acak, algoritma k-means cenderung menghasilkan klaster yang berbeda.…”

Section: Pendahuluanunclassified

See 1 more Smart Citation

Penerapan Algoritma Pillar Untuk Inisialisasi Titik Pusat K-Means Klaster Dinamis

Seputra

Wijaya

2020

JTIIK

View full text Add to dashboard Cite

<p>K-Means merupakan algoritma yang digunakan untuk melakukan pengklasteran data. Namun, k-means memiliki<br />masalah dalam sensitivitas penentuan partisi awal jumlah klaster. Penelitian terkait menyatakan algoritma k-means tergantung pada penentuan titik pusat klaster awal. Pemilihan pusat klaster awal secara acak cenderung menghasilkan klaster yang berbeda. Sehingga untuk menentukan klaster terbaik harus dilakukan dengan memperhatikan nilai Sum Sequare Error yang terkecil. Untuk mengatasi permasalahan tersebut, penentuan klaster dilakukan dengan menggunakan algoritma pillar. Algoritma pillar menentukan titik pusat klaster dengan memilih data dengan nilai euclidean paling jauh dari titik pusat klaster. Namun pemilihan titik klaster tetap memperhatikan kemungkinan data outlier. Pengujian dilakukan dengan menetapkan satu buah klaster awal sebagai inisialisasi skaligus sebagai klaster pembanding untuk menentukan kualitas klaster berikutnya. Penelitian ini menggunakan data set ruspini dan iris. Untuk data ruspini terdiri dari 76 data set, sedangkan data iris terdiri dari 150 data set. Klaster Pillar memiliki nilai Sum Sequere Error, Variance Cluster, dan Davies yang lebih kecil dibandingkan klaster dinamis pada data set ruspini. Nilai tersebut secara berurutan untuk algoritma pillar adalah 0.28, 0.11, 7.30, 5.88. Untuk data set iris nilai Sum Square Error lebih tinggi dibandingkan dengan klaster dinamis yaitu 0.34. Sedangkan algoritma klaster dinamis memiliki nilai 0.32. Hal tersebut disebahkan penentuan data outlier pada iris data set yang tidak akurat. Ketidakakurantan tersebut berasal dari data yang bersifat multivariat, sehingga memungkinkan data outlier menjadi centroid awal klaster. Sehingga jika dilihat dari nilai validitas SSE, algoritma pillar k-means klaster dinamis masih kurang bekerja optimal dibandingkan dengan algoritma k-means klaster dinamis.</p>

show abstract

Section: Pendahuluanunclassified

Penerapan Algoritma Pillar Untuk Inisialisasi Titik Pusat K-Means Klaster Dinamis

Seputra

Wijaya

2020

JTIIK

View full text Add to dashboard Cite

show abstract

“…Normally the risk factor is identified by machine learning and deep learning algorithm. This work focuses on the accuracy measures in point, we utilize the ELM algorithm with CS-SVM [4] for better optimization and accuracy. For Structured data, Naïve Bayes (NB), CS-SVM, Decision tree (DT) Machine learning algorithm is employed to find the risk of fatal disease.…”

Section: A Risk Factor Predictionmentioning

confidence: 99%

Big Data for Health Care Analytics using Extreme Machine Learning Based on Map Reduce

Karuppan,

Nithya,

Ondimuthu

2020

IJEAT

View full text Add to dashboard Cite

A large volume of datasets is available in various fields that are stored to be somewhere which is called big data. Big Data healthcare has clinical data set of every patient records in huge amount and they are maintained by Electronic Health Records (EHR). More than 80 % of clinical data is the unstructured format and reposit in hundreds of forms. The challenges and demand for data storage, analysis is to handling large datasets in terms of efficiency and scalability. Hadoop Map reduces framework uses big data to store and operate any kinds of data speedily. It is not solely meant for storage system however conjointly a platform for information storage moreover as processing. It is scalable and fault-tolerant to the systems. Also, the prediction of the data sets is handled by machine learning algorithm. This work focuses on the Extreme Machine Learning algorithm (ELM) that can utilize the optimized way of finding a solution to find disease risk prediction by combining ELM with Cuckoo Search optimization-based Support Vector Machine (CS-SVM). The proposed work also considers the scalability and accuracy of big data models, thus the proposed algorithm greatly achieves the computing work and got good results in performance of both veracity and efficiency.

show abstract

“…Moreover, dynamic replanning of Voronoi diagrams is used for persistent task [31]. K-means clustering is also a very effective method in data science which can promote the workload balance in subregions [32].…”

Section: Related Workmentioning

confidence: 99%

Dynamic Control Scheme of Multiswarm Persistent Surveillance in a Changing Environment

Jing

Wang

et al. 2019

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

The persistent surveillance problem has been proved to be an NP hard problem for multiple unmanned aerial vehicle systems (UAVs). However, most studies in multiple UAV control focus on control cooperative path planning in a single swarm, while dynamic deployment of a multiswarm system is neglected. This paper proposes a collective control scheme to drive a multiswarm UAVs system to spread out over a time-sensible environment to provide persistent adaptive sensor coverage in event-related surveillance scenarios. We design the digital turf model to approximate the mixture information of mission requirements and surveillance reward. Moreover, we design a data clustering-based algorithm for the dynamic assignment of UAV swarms, which can promote workload balance, while also allowing real-time response to emergencies. Finally, we evaluate the proposed architecture by means of simulation and find that our method is superior to the conventional control strategy in terms of detection efficiency and subswarm equilibrium degree.

show abstract

An Improved K-means Algorithm based on Mapreduce and Grid

Abstract: The traditional K-means

Cited by 18 publications

References 19 publications

Penerapan Algoritma Pillar Untuk Inisialisasi Titik Pusat K-Means Klaster Dinamis

Penerapan Algoritma Pillar Untuk Inisialisasi Titik Pusat K-Means Klaster Dinamis

Big Data for Health Care Analytics using Extreme Machine Learning Based on Map Reduce

Dynamic Control Scheme of Multiswarm Persistent Surveillance in a Changing Environment

Contact Info

Product

Resources

About