NEVEEN IBRAHIM MOHAMED GHALI

nehal hamdy al-banhawy

01/12/2020

Handwritten signature identification and verification has become an active area of research in recent years. Handwritten signature identification systems are used for identifying the user among all users enrolled in the system while handwritten signature verification systems are used for authenticating a user by comparing a specific signature with his signature that is stored in the system. This paper presents a review for commonly used methods for pre-processing, feature extraction and classification techniques in signature identification and verification systems, in addition to a comparison between the systems implemented in the literature for identification techniques and verification techniques in online and offline systems with taking into consideration the datasets used and results for each system.

NEVEEN IBRAHIM MOHAMED GHALI

SHEREEN A. EL-AAL, AHMED A. A. GAD-ELRAB, AFAF A. S. ZAGHROUT

01/09/2020

Recently, Internet of Things (IoTs) influences every aspect of human daily lives through intelligent systems as healthcare, traffic management, and smart building. These IoTs systems depend on contextualization of collecting data through context aware system to gain knowledge by using context reasoning. context reasoning is a way for deducing knowledge and providing better understanding of the collected raw data. Context reasoning is commonly carried out at the cloud due to its high processing capabilities. However, the main challenges of using cloud are high latency time and resource consumption. To meet these challenges, Fog computing is proposed as an intermediate layer between the IoTs devices and the cloud layer to comply IoTs requirements of latency time reduction and resource consumption by deploying services to the fog layer. In this paper a new context reasoning model is proposed based on three previously defined Deep Learning (DL) models which are GoogleNet, ResNet101 and DenseNet201, the results obtained in three cases are compared in cloud and cloud/fog environments. The conducted simulation experiments with fog showed that the proposed cloud/fog model can reduce the time delay, execution time, and energy consumption with good classification accuracy which is up to 96%. These reduction values are 4%, 10%, and 94%, respectively, less than values by using cloud layer.

NEVEEN IBRAHIM MOHAMED GHALI

Asaad Ahmed,Shereen A. El-aal,Afaf A. S. Zaghrout

01/02/2020

Internet of Things (loTs) is the future of ubiquitous and personalized intelligent service delivery. It depends on installing intelligent sensors to sense and control physical environment to generate enormous amount of data with various data types. Context aware computing is employed for transforming these sensor data into knowledge through three stages: collection, modeling and reasoning. In context modeling, raw data represents in according meaningful manner statically. Furthermore, with growth of IoTs live applications, static modeling is not convenient because of changing context data structure overtime. The work in this paper is dedicated to propose a new dynamic approach for context modeling based on genetic algorithm and satisfaction factor. In addition, flexibility indicator property and context based are defined to measure the performance of the proposed approach

NEVEEN IBRAHIM MOHAMED GHALI

Ahmed. A. A. Gad-Elrab , Shereen A. El-aal , Afaf A. S. Zaghrout

01/12/2019

Internet of Things (IoTs) enables entities every day to communicate and collaborate with each other for providing information, data and services to inhabitants and users. IoTs consists of a large number of smart devices that can generate immense amount of data with different types. These sensors raw data needs to be modeled in a certain structure before filtering and processing to provision context information. This process is called context modeling. Context modeling provides definition of how context data are structured and maintained through context aware system. However, employing model for every context type through context aware application is static and is specified by the application developer. The main problem in IoTs is that the structure of context data changes overtime, therefore static modeling cannot be adaptable for modeling these changes. In this paper, a new dynamic approach for context modeling based on genetic algorithm and satisfaction factor is proposed. Firstly, the proposed approach uses genetic algorithm to find the best matching between a set of contexts and a set of available context models. Secondly, it uses a satisfaction factor to calculate the satisfaction degree for each context with each available context model and select the context model with high satisfaction degree as the structure model of this context, dynamically. In addition, flexibility indicator property and context based are defined to measure the performance of the proposed approach. The results of conducted simulations show that the proposed approach achieves higher performance than static approach for context modeling.

NEVEEN IBRAHIM MOHAMED GHALI

Amany A. Naem

01/12/2019

Antlion Optimization (ALO) is one of the latest population based optimization methods that proved its good performance in a variety of applications. The ALO algorithm copies the hunting mechanism of antlions to ants in nature. Community detection in social networks is conclusive to understanding the concepts of the networks. Identifying network communities can be viewed as a problem of clustering a set of nodes into communities. k-median clustering is one of the popular techniques that has been applied in clustering. The problem of clustering network can be formalized as an optimization problem where a qualitatively objective function that captures the intuition of a cluster as a set of nodes with better in ternal connectivity than external connectivity is selected to be optimized. In this paper, a mixture antlion optimization and k-median for solving the community detection problem is proposed and named as K-median Modularity ALO. Experimental results which are applied on real life networks show the ability of the mixture antlion optimization and k-median to detect successfully an optimized community structure based on putting the modularity as an objective function.

NEVEEN IBRAHIM MOHAMED GHALI

Yomna M. Elbarawy,Rania Salah El-Sayed

01/10/2019

Facial expressions are undoubtedly the best way to express human attitude which is crucial in social communications. This paper gives attention for exploring the human sentimental state in thermal images through Facial Expression Recognition (FER) by utilizing Convolutional Neural Network (CNN). Most traditional approaches largely depend on feature extraction and classification methods with a big pre-processing level but CNN as a type of deep learning methods, can automatically learn and distinguish influential features from the raw data of images through its own multiple layers. Obtained experimental results over the IRIS database show that the use of CNN architecture has a 96.7% recognition rate which is high compared with Neural Networks (NN), Autoencoder (AE) and other traditional recognition methods as Local Standard Deviation (LSD), Principle Component Analysis (PCA) and K-Nearest Neighbor (KNN)

NEVEEN IBRAHIM MOHAMED GHALI

Yomna M. Elbarawy, Rania Salah El-Sayed

01/12/2018

Emotional reactions are the best way to express human attitude and thermal imaging mainly used to utilize detection of temperature variations as in detecting spatial and temporal variation in the water status of grapevine. By merging the two facts this paper presents the Discrete Cosine Transform (DCT) with Local Entropy (LE) and Local Standard Deviation (LSD) features as an efficient filters for investigating human emotional state in thermal images. Two well known classifiers, K-Nearest Neighbor (KNN) and Support Vector Machine (SVM) were combined with the earlier features and applied over a database with variant illumination, as well as occlusion by glasses and poses to generate a recognition model of facial expressions in thermal images. KNN based on DCT and LE gives the best accuracy compared with other classifier and features results.

NEVEEN IBRAHIM MOHAMED GHALI

Amany A. Naem,Afaf A. Saleh

01/12/2018

Spam emails are not necessary, though they are harmful as they include viruses and spyware, so there is an emerging need for detecting spam emails. Several methods for detecting spam emails were suggested based on the methods of machine learning, which were submitted to reduce non-relevant emails and get results of high precision for spam email classification. In this work, a new predictive method is submitted based on antlion optimization (ALO) and boosting termed as ALO-Boosting for solving spam emails problem. ALO is a computational model imitates the preying technicality of antlions to ants in the life cycle. Where ALO was utilized to modify the actual place of the population in the separate seeking area, thus obtaining the optimum feature subset for the better classification submit based on boosting classifier. Boosting classifier is a classification algorithm that points to a group of algorithms which modifies soft learners into powerful learners. The proposed procedure is compared against support vector machine (SVM), k-nearest neighbours algorithm (KNN), and bootstrap aggregating (Bagging) on spam email datasets in a set of implementation measures. The experimental outcomes show the ability of the proposed method to successfully detect optimum features with the smallest value of selected features and a high precision of measures for spam email classification based on boosting classifier.