No abstract

No abstract

No abstract

Human activity recognition is influential subject in different fields of human daily life especially in the mobile health. As the smartphone becomes an integrated part of human daily life which has the ability of complex computation, internet connection and also contains a large number of hardware sensors, encourage implementation of the human activity recognition system. Most of the works done in this field imposed the restriction of firmly fixing the smartphone in a certain position on the human body, together with machine learning mechanism, to facilitate the process of classifying human activities from the smartphone sensors raw data. To overcome this restriction, the proposed approach incorporated a smartwatch, fixed on the human ankle, together with smartphone freely carried by the user. The use of smartwatch assisted in providing distinctly separable signal variation from the smartwatch accelerometer and gyroscope sensors raw data which in turn facilitated the use of a threshold-based mechanism to classify 20 various human activities. Furthermore, this work provides a service for remotely real-time monitoring of the user human activities the system is tested with different subjects and achieved an accuracy of 97.5%. .

No abstract

Steganography is a technique in which a person hides information in digital media. The message sent by this technique is so secret that other people cannot even imagine the information’s existence. This article entails developing a mechanism for communicating one-on-one with individuals by concealing information from the rest of the group. Based on their availability, digital images are the most suited components for use as transmitters when compared to other objects available on the internet. The proposed technique encrypts a message within an image. There are several steganographic techniques for hiding hidden information in photographs, some of which are more difficult than others, and each has its strengths and weaknesses. The encryption mechanism employed may have different requirements depending on the application. For example, certain applications may require complete invisibility of the key information, while others may require the concealment of a larger secret message. In this research, we proposed a technique that converts plain text to ciphertext and encodes it in a picture using up to the four least significant bit (LSB) based on a hash function. The LSBs of the image pixel values are used to substitute pieces of text. Human eyes cannot predict the variation between the initial Image and the resulting image since only the LSBs are modified. The proposed technique is compared with state-of-the-art techniques. The results reveal that the proposed technique outperforms the existing techniques concerning security and efficiency with adequate MSE and PSNR.

In protein fold recognition problem an effort is made to assign a fold to given proteins, this is of practical importance and has diverse application in the field of bioinformatics such as the discovery of new drugs, the individual implication of amino acid in a protein and bringing improvement in a specific protein function. In this paper, we have studied various machine learning techniques for protein fold recognition problem, and compared Support Vector Machine (SVM) with Radial Basis Function (RBF) kernel and Multilayer Perceptron (MLP) on a number of measures like the recognition accuracy of protein fold, the 10-fold cross validation accuracies and Kappa statistics. These techniques are applied to the well known Structural Classification of Proteins (SCOP) dataset in extensive experimentations. In this study Multilayer Perceptron (MLP) shows better accuracy on single protein feature (C, S, H, P, V, Z) of the SCOP dataset as compared to Support Vector Machine (SVM). A plausible reason of the better performance of MLP is that it uses all the available data for classification where as the SVM model cannot exploit all the available data.

In this chapter, the authors propose a multi-objective solution to the problem by using multi-objective particle swarm optimization (MOPSO) algorithm to optimize the number of clusters in a sensor network in order to provide an energy-efficient solution. The proposed algorithm considers the ideal degree of nodes and battery power consumption of the sensor nodes. The main advantage of the proposed method is that it provides a set of solutions at a time. The results of the proposed approach were compared with two other well-known clustering techniques: WCA and CLPSO-based clustering. Extensive simulations were performed to show that the proposed approach is an effective approach for clustering in WSN environments and performs better than the other two approaches.