Open AccessObject Detection using K-Means Clustering – A Research
Author(s) -
Madhura P. Divakara,
Keerthi V. Trimal,
Adithi Krishnan,
V Karthik
Publication year - 2019
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.b1160.0882s819
Subject(s) - cluster analysis , correlation clustering , computer science , cure data clustering algorithm , k medians clustering , single linkage clustering , cluster (spacecraft) , fuzzy clustering , object (grammar) , focus (optics) , set (abstract data type) , k means clustering , task (project management) , path (computing) , artificial intelligence , group (periodic table) , determining the number of clusters in a data set , clustering high dimensional data , data mining , pattern recognition (psychology) , engineering , physics , systems engineering , optics , programming language , quantum mechanics
Clustering is an unsupervised machine learning technique and the task is to group a set of objects in such a way that objects in the same group are more similar to each other than those in other groups. There are different clustering techniques, each with its own advantages and disadvantages. The K-means clustering algorithm is our main focus in this paper. K-means is mostly used when there is a large number of unlabeled data. The difficulties in the path of K-means clustering are a) Different initial points can lead to different final clusters. b) It does not work with clusters of different sizes and densities. c) With the global cluster, it does not work well and is difficult to determine K Value. Our main aim is to try and modify the K-means clustering algorithm to get rid of the above-mentioned drawbacks.