This paper presents a data clustering approach using modified K-Means algorithm based on the improvement of the sensitivity of initial center (seed point) of clusters. This algorithm partitions the whole space into different segments and calculates the frequency of data point in each segment. The segment which shows maximum frequency of data point will have the maximum probability to contain the centroid of cluster. The number of cluster's centroid (k) will be provided by the user in the same manner like the traditional K-mean algorithm and the number of division will be k∗k (‘k’ vertically as well as ‘k’ horizontally). If the highest frequency of data point is same in different segments and the upper bound of segment crosses the threshold ‘k’ then merging of different segments become mandatory and then take the highest k segment for calculating the initial centroid (seed point) of clusters. In this paper we also define a threshold distance for each cluster's centroid to compare the distance between data point and cluster's centroid with this threshold distance through which we can minimize the computational effort during calculation of distance between data point and cluster's centroid. It is shown that how the modified k-mean algorithm will decrease the complexity & the effort of numerical calculation, maintaining the easiness of implementing the k-mean algorithm. It assigns the data point to their appropriate class or cluster more effectively.

Data clustering with modified K-means algorithm

In this paper, a novel electroencephalographic (EEG)-based method is introduced for the quantification of brain-electrical connectivity changes over a longitudinal evaluation of mild cognitive impaired (MCI) subjects. In the proposed method, a dissimilarity matrix is constructed by estimating the coupling strength between every pair of EEG signals, Hierarchical clustering is then applied to group the related electrodes according to the dissimilarity estimated on pairs of EEG recordings. Subsequently, the connectivity density of the electrodes network is calculated. The technique was tested over two different coupling strength descriptors: wavelet coherence (WC) and permutation Jaccard distance (PJD), a novel metric of coupling strength between time series introduced in this paper. Twenty-five MCI patients were enrolled within a follow-up program that consisted of two successive evaluations, at time T0 and at time T1, three months later. At T1, four subjects were diagnosed to have converted to Alzheimer’s Disease (AD). When applying the PJD-based method, the converted patients exhibited a significantly increased PJD ( $p ), i.e., a reduced overall coupling strength, specifically in delta and theta bands and in the overall range (0.5–32 Hz). In addition, in contrast to stable MCI patients, converted patients exhibited a network density reduction in every subband (delta, theta, alpha, and beta). When WC was used as coupling strength descriptor, the method resulted in a less sensitive and specific outcome. The proposed method, mixing nonlinear analysis to a machine learning approach, appears to provide an objective evaluation of the connectivity density modifications associated to the MCI-AD conversion, just processing noninvasive EEG signals.

Permutation Jaccard Distance-Based Hierarchical Clustering to Estimate EEG Network Density Modifications in MCI Subjects

This paper evaluates performance of two clustering algorithms, namely k-means and mini batch k-means, in the Android malware detection. Network traffic generated by the Android applications, normal and malicious, is analyzed for detection purpose. We have used MalGenome data sample for this work to build the dataset. We chose 800 samples out of 1260 Android malware samples. In addition, we collected numerous normal applications from the official Android market. The results show that mini batch k-means algorithm performs better than k-means algorithm in the Android malware detection.

Comparative study of k-means and mini batch k-means clustering algorithms in android malware detection using network traffic analysis

In the present scenario, humans are surrounded by numerous diseases which affect their lifestyles. One such condition is glaucoma which originates due to sudden stress in the eye, generally termed as intraocular pressure culminating in the affliction of the optic nerve and permanent loss of vision. It is the second highest characteristic root of blindness. The optic nerve is liable for a transference of visual details to the brain from retina leading one to locate the exterior world. Clinical diagnosis of glaucoma includes an examination of the optic cup, optic disc and measurement of intraocular pressure by tonometer, pachymetry, etc. All these tasks are manual and time-consuming hence, a computer-aided diagnosis (CAD) system is needed for evaluation of glaucoma which can be done by analysis of optic disc and optic cup present in retinal fundus images using image processing approaches. Here, optic disc represents the entrance of optic nerve and site from where retinal nerve cells approach towards each other while, the optic cup is the centric depression of irregular size residing on the disc. In this work, a Level Set Based Adaptively Regularized Kernel-Based Intuitionistic Fuzzy C means (LARKIFCM) based approach is proposed for segmentation of optic disc and optic cup in retinal fundus images with improved performance for accurate diagnosis.

Optic disc and optic cup segmentation from retinal images using hybrid approach

Lung cancer is the major cause of death among patients with cancer throughout the world. The main symptom that indicate the lung cancer is the presence of lung nodules. This work proposes a methodology to classify lung nodule and non-nodule using texture features. The state-of-art of the presented work are the adaption of the Artificial Crawlers and Rose Diagram techniques for representing patterns over 3D images. Several information are extracted based on the texture behavior of these methods, allowing the correct classification of lung nodules candidates using Support Vector. Objective: This work proposes a methodology to classify lung nodule candidates and non-nodule candidates based on computed tomography (CT) images. Methodology: The Lung Image Database Consortium (LIDC-IDRI) image database is employed for our tests. Three techniques are employed to extract texture measurements. The first technique is artificial crawlers (ACs), an artificial life algorithm. The second technique uses the rose diagram (RD) to extract directional measurements. The third technique is a hybrid model that combines texture measurements from artificial crawlers and the rose diagram. The support vector machine (SVM) classifier with a radial basis kernel is employed. Results: In the testing stage, we used 833 scans from the LIDC-IDRI database. For the application of the methodology, we decided to divide the whole database into two groups, training and testing. We used partitions of training and testing of 20/80%, 40/60%, 60/40% and 80/20%. The division was repeated 5 times at random. We reached a mean accuracy (mACC) of 94.30%, a mean sensitivity (mSEN) of 91.86%, a mean specificity (mSPC) of 94.78%, a coefficient of accuracy variance (CAv) of 1.61% and a mean area under the receiver operating characteristic (mROC) curves of 0.922. Conclusion: Lung cancer has the highest mortality rate and one of the smallest survival rates after diagnosis. An early diagnosis increases the survival chance of patients. The proposed methodology is a useful tool for specialists in the detection of nodules. We believe we contribute for the expert system field because 1) the adaption of the Artificial Crawlers and Rose Diagram methods as 3D texture descriptors is innovative and contains great potential; 2) we adapted and developed measurements from the 3D texture descriptors; and 3) the simplicity and discriminative power of the methodology can be extended to applications based on images with other contexts.

Lung nodule classification using artificial crawlers, directional texture and support vector machine

Document Clustering is a widely studied problem in Text Categorization. It is the process of partitioning or grouping a given set of documents into disjoint clusters where documents in the same cluster are similar. K-means, one of the simplest unsupervised learning algorithms, solves the well known clustering problem following a simple and easy way to classify a given data set through a certain number of clusters (assume k clusters) fixed a priori. The main idea is to define k centroids, one for each cluster. This clustering algorithm uses an iterative procedure which converges to one of numerous local minima. We have found that these iterative techniques are especially sensitive to initial starting conditions of the centroid of each cluster and the more the distance among the cluster centroid the better the clustering performance. In simple K-means algorithm the way to initialize the centroid is not specified and one popular way to start is to randomly choose k points of the samples as k centroids but this process does not guarantee to choose the maximum dissimilar documents as the centroid point for k-cluster. In this paper we proposed a modified k-means algorithm which uses Jaccard distance measure for computing the most dissimilar k documents as centroids for k clusters. Our experimental results demonstrate that our proposed K-means algorithm with Jaccard distance measure for computing the centroid improves the clustering performance of the simple K-means algorithm.

An efficient k-means algorithm integrated with Jaccard distance measure for document clustering

Wireless testing depends on the wireless test environments where test environment refers to the setup or environment into which the device being tested and the test equipment are placed. Four primary types of test environments are Faraday Cages, RF Chambers, Cabled and Open air. Most wireless testing is conducted in indoor environment which is veritable RF jungles, especially when we consider the ISM bands. It is important to understand the strengths and weaknesses of the range of wireless test environments because the impact of the wireless environment is often underestimated. But in reality wireless environment is responsible for producing different results of an experiment. We tried to find out a perfect test environment for wireless testing specially when they venture into the RF jungle. Our aim is to find out a testbed which provide better control over the RF environment to experimenters.

Efficient testing environment for wireless networking experiments

This research paper concerns with the effectiveness of IR-UWB signals can be analyzed using the recursive configuration of predictable TR and standard TR systems. This can be achieved by using IEEE 802.15.4a specification which is contrasted with predictable TR using BER which is thought to be quasi-static channel. However in case of standard TR system, its progression could be little bit changed. The matlab simulation evaluates the performance of the projected system is better than the performance of the traditional TR systems.

Performance analysis and recognition of IR-UWB signals using TR system

Quality of Service support is a major challenge in VoIP. So much research has been devoted recently on maintaining an acceptable Quality-of-Service (QoS) for the transmission of real-time multimedia streams. In personal communication human speech processing is one of the most important procedure. Because if a low bit-rate codec is used in voice over IP (VoIP) to achieve a high compression ratio, the original waveform can be significantly distorted. So this transmission impairment may decrease the quality of the speech, as a result we have error in the resulting speech. These new potential sources of signal distortion present significant challenges for objectively measuring speech quality. To overcome this problem we need to find an efficient speech enhancement algorithm. In this dissertation, adaptive gain equalizer algorithm is used for the accomplishment of a time domain technique for the enhancement of given speech. Here the concept of Analog Electronic Circuit may be very useful as our aim to use a printed board circuit where ordinary analog electronic equipments are used. The projected clarification helps commencing extreme level of system bandwidth which does not need to quantizes as well as digitalizes information and put away extra resourceful power while go up against to various digital signal processor sup-ported clarifications or resolutions. Tests using our proposed algorithm prove the overall enhancement of speech quality on voice over internet protocol.

Notice of Violation of IEEE Publication Principles Adaptive gain equalizer algorithm using analog electronic circuit for speech enhancement in voice over internet protocol

Packet loss is a common challenge to provide efficient Quality of Service support in VoIP. Present VoIP architecture does not guaran- tee a minimum amount of packet loss and a mini- mum delay of packet transmission required for voice communication. The recent work has shown that in voice transmission of all the packets are not of same importance or priority. Some seg- ments of the signal are essential to the speech quality while others are not so much essential and this gives the concept of prioritizing the informa- tive packets of the voiced stream which provides good voice quality and also ensures the efficient use of the network resources. To provide efficient QoS support selective packet marking procedure is better than using "best effort" principle. We have proposed to use SPB-MARK scheme for prioritization of voiced packets. Again using packet loss concealment algorithm (PLC), the quality of service for voice over IP can be im- proved but the PLC techniques described in the standards ANSI TI.521 and ITU-T Rec. G.711 fail to reconstruct the lost packet if it is the first packet in the voiced stream as no previous pack- ets are available for reconstruction of that packet. For this, we propose a special buffer to store the information of first three voiced packets for the reconstruction of the lost packets. Tests on recon- structed packets those are lost have proved that the proposed method improves the reconstructed speech quality significantly.

Mushfeq-Us-Saleheen Shameem

Papers

An efficient k-means algorithm integrated with Jaccard distance measure for document clustering

Efficient testing environment for wireless networking experiments

Performance analysis and recognition of IR-UWB signals using TR system

Notice of Violation of IEEE Publication Principles Adaptive gain equalizer algorithm using analog electronic circuit for speech enhancement in voice over internet protocol

Design of an efficient Receiver based Packet Loss Concealment technique for Voice over Internet Protocol Using Speech Property-Based Marking Scheme