ICTACT Journal on Soft Computing

Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Fuzzy Clustering Algorithms-Comparative Studies for Noisy Speech Signals


Affiliations
1 Department of Information Science and Engineering, JSS Science and Technology University, India
2 Department of Computer Science and Engineering, JSS Science and Technology University, India
3 Department of Computer Science and Engineering, JSS Academy of Technical Education, India
     

   Subscribe/Renew Journal


In the area of speech signal processing and recognition, application of soft computing techniques is one of the prominent techniques for clustering the overlapping data. Kernel FCM technique is one of the efficient method to cluster the data by computing the cluster centroids. This paper presents and compares the most important clustering techniques like k-means, Fuzzy C means and Kernel Fuzzy C Means algorithms for clustering noisy speech signals. The clustering performances of these techniques are tabulated for homogeneous and heterogeneous speech data sets. This paper highlights the importance of KFCM algorithm for clustering the overlapping data. It also demonstrates the computation time and recognition accuracies of each technique. Our study identifies the KFCM technique performs better than k-means and FCM techniques.

Keywords

Additive Noise, Clustering, Convolved Noise, Fuzzy C Means (FCM), Heterogeneous Data, Homogeneous Data, K-Means, Kernel Fuzzy C Means (KFCM), Principal Component Analysis (PCA), Validity Measures.
Subscription Login to verify subscription
User
Notifications
Font Size


  • Fuzzy Clustering Algorithms-Comparative Studies for Noisy Speech Signals

Abstract Views: 382  |  PDF Views: 0

Authors

H. Y. Vani
Department of Information Science and Engineering, JSS Science and Technology University, India
M. A. Anusuya
Department of Computer Science and Engineering, JSS Science and Technology University, India
M. L. Chayadevi
Department of Computer Science and Engineering, JSS Academy of Technical Education, India

Abstract


In the area of speech signal processing and recognition, application of soft computing techniques is one of the prominent techniques for clustering the overlapping data. Kernel FCM technique is one of the efficient method to cluster the data by computing the cluster centroids. This paper presents and compares the most important clustering techniques like k-means, Fuzzy C means and Kernel Fuzzy C Means algorithms for clustering noisy speech signals. The clustering performances of these techniques are tabulated for homogeneous and heterogeneous speech data sets. This paper highlights the importance of KFCM algorithm for clustering the overlapping data. It also demonstrates the computation time and recognition accuracies of each technique. Our study identifies the KFCM technique performs better than k-means and FCM techniques.

Keywords


Additive Noise, Clustering, Convolved Noise, Fuzzy C Means (FCM), Heterogeneous Data, Homogeneous Data, K-Means, Kernel Fuzzy C Means (KFCM), Principal Component Analysis (PCA), Validity Measures.

References