Search

Reliance on deep learning techniques has become an important trend in several science domains including biological science, due to its proven efficiency in manipulating big data that are often characterized by their non-linear processes and complicated relationships. In this study, Convolutional Neural Networks (CNN) has been recruited, as one of the deep learning techniques, to be used in classifying and predicting the biological activities of the essential oil-producing plant/s through their chemical compositions. The model is established based on the available chemical composition’s information of a set of endemic Egyptian plants and their biological activities. Another type of machine learning algorithms, Multiclass Neural Network (MNN), has been applied on the same Essential Oils (EO) dataset. This aims to fairly evaluate the performance of the proposed CNN model. The recorded accuracy in the testing process for both CNN and MNN is 98.13% and 81.88%, respectively. Finally, the CNN technique has been adopted as a reliable model for classifying and predicting the bioactivities of the Egyptian EO-containing plants. The overall accuracy for the final prediction process is reported as approximately 97%. Hereby, the proposed deep learning model could be utilized as an efficient model in predicting the bioactivities of, at least Egyptian, EOs-producing plants.

The volatile nature of Cloud user’s needs is considered a worthy challenge in formulating the Service Level Agreement (SLA) negotiation. The sterility of the standard SLA templates in handling the ...

Improving the performance of e-learning services to provide a scalable and effective e-learning system is a big challenge for the educational organizations. The e-learning system faces many challenges in pedagogical services (e.g. learning design and learning content problems), and technical services (e.g. resource provisioning and financial cost). This article presents an e-learning environment based on cloud computing (PCLE) in attempt to enhance the e-learning services' by customizing the contents of the course's material depending on the students' knowledge, experiences, and requirements. Also, the system focuses on supporting and achieving reusability, interoperability, adaptation and personalization in order to overcome the passive role of the student and transfer it into an interactive and useful participation. On the other hand, PCLE is built over the cloud computing environment trying to overcome the traditional web-hosting challenges as scalability, availability, and cost.

Throughout the years Machine Learning (ML) has increased a lot of consideration on ordinary products as search, filters, recognition and recently genomics. Various strategies incorporate sophisticated artificial neural system designs and are all known as applications of Deep Learning (DL). These days, deep learning could be a current and a fortifying field of machine learning. Deep Learning models have fair been shown prepared for both enhancing data encoding simplicity and prescient design execution over elective methodologies. Also deep learning techniques have been shown to reflect and learn unsurprising relationships in various different types of data and to guarantee that the future of genomics research and precise medicine applications will change. DL applications in genomic field is rapidly developed.This review presents a clarification for machine learning and deep learning methods utilized in genomics. And the main goal is to show a detailed comprehensive overview on the available ML and DL techniques used in genomics.

Real-time systems are intensively used nowadays. Scheduling algorithms are very important to manage the scheduling of tasks in real-time systems. In this paper we give an overview on the real-time scheduling techniques for uniprocessors and multiprocessors, then we present a comparison between the multiprocessor scheduling algorithms which are classified into partitioning and global scheduling algorithms. The results achieved from the comparison have showed that the parameters such as makespan, waiting time, missed deadlines and task preemptions are better in performance when using partitioned scheduling algorithms than that for global scheduling algorithms when the number of tasks is small. While when the number of tasks increased, it gives more better performance when using global scheduling algorithms than partitioned scheduling algorithms using all parameters except for some of the algorithms gave low performance in missed deadlines, they give a high number of missed deadlines. But in total when the number of tasks increased the global scheduling algorithms gives better performance than partitioned scheduling algorithms.

Cloud computing lineaments enable the users to deal with resources and get benefit from the services in a distinct economic manner. One of the vital cloud’s services is storing the users’ data on a remote location on a cloud’s storage resource, which means that this storage has to be characterized by a high level of security and safety, for instance, to ensure the data integrity and preserve its confidentiality. In the existing cloud environments, these security issues are handled by a third party; neither user nor provider, known as the Cloud Third Party (CTP). Despite the evident important role of CTP, it may adversely affect the security and becomes untrustworthy insider employee (i.e., malicious insider). In this paper, we present an automated framework for enhancing the cloud security “Self-Protecting of Data Auditing (SPDA)” based on the concept of autonomic computing. This framework is divided into three phases in order to handle the two significant security issues: the integrity and the confidentiality. The first stage concerns with encrypting the user’s data automatically to maintain its confidentiality. The second phase is for automated data auditing in order to keep the data integrity. And, the last one is used to automatically decrypt the data to its original correct form when the user requests to retrieve it.

One of the most significant research fields in image processing is image fusion that means combining two or more images with different characteristics in order to integrate relevant information from the source images into a rich informative image. Generally, the fusion methods are categorized into two domains; the spatial domain methods that deal directly with the pixel value in the input images to obtain the fused image, and the frequency domain methods in which the source images are first transformed into the frequency domain by computing the Fourier Transform of the source images, where the fusion operations are all applied on the Fourier transform and then applying inverse Fourier transform to obtain the fused image. In this paper, a comparison among some of the recent image fusion techniques in both the spatial and transform domains using a medical image dataset of two different medical imaging modalities (CT and MRI). The techniques that have been used in the experiment include PCA and Simple Average in the spatial domain and DWT, SWT, LP, DCT, and LP+DCT in the transform domain. The performance parameters used in the evaluation step were: Standard Deviation (SD), and Peak Signal to Noise Ratio (PSNR). Experimental results illustrate some features of the spatial domain methods having a better peak signal to noise ratio and thus keep the similarity between the fused image and the source images, reduce image distortion, as well as they are simple and easy to implement. On the other hand, the multiresolution methods give better information and high contrast and enhance spectral information.

Within medical imaging, there are various modalities of medical images like CT, X-rays, MRI and other modalities that provide information about a human body in different ways. Each modality has distinctive characteristics that provide various sources of information. Therefore, there are some problems like image comparison such as CT/PET, CT /MRI, and MRI/ PET were usually meet by the clinical treatment and diagnosis. Hence the need to combine the different images' information and this process is known as ‘medical image fusion’. In this paper, two techniques for the ‘medical image fusion’ are introduced. The first proposed fusion technique is the combination of the fast filtering with the discrete wavelet transform ‘DWT’ methods for overcoming the low spatial resolution fused image provided by DWT and preserve the source images' salient features. Where we used the fast filtering method procedures for combining the corresponding ‘low-frequency coefficients’ to maintain the ‘salient features’ of the initial images, and the maximum rule with the high-frequency coefficients which lead getting better the resultant image contrast. The second proposed technique is the combination of fast filtering with stationary wavelet transform (SWT) methods, where ‘SWT’ has the shift-invariant property which enables to overcome the shift-variance DWT's drawback. The performance of the fused output is tested and compared with five of the common fusion methods like the Gradient pyramid, Contrast pyramid, DWT, Fast Filtering, and SWT techniques, using performance parameters: E, SNR, SD, and PSNR.

Cloud Computing is, by nature, multi-tenant, complex, large-scale, and heterogeneous distributed systems. Thus respectively, its processes and strategies need to be automated and integrated. one of the essential processes in the Cloud computing system is negotiating the service level agreement which always has to be elastic and flexible in handling and translating the user services’ requirements. This paper aims at presenting an automated SLA negotiation framework based on autonomous and flexible agents and multi agent systems based on the autonomic computing features as suitable tools for self-detection of failures and self-monitoring for the Cloud operations and services.

To increase our understanding of the interactions between the different cellular and intercellular processes, it is imperative to construct accurate biological models. An important class of these models are usually used to predict the time evolution of the corresponding chemical species over a specified period of time. Stochastic and coloured stochastic Petri nets can be adopted to construct graphical models that can be executed to simulate the firing of a set of reactions in a well-mixed chemical system. For the purpose of this specific application, (coloured) stochastic Petri nets employ internally the stochastic simulation algorithm (SSA) to perform the simulation. Although the result of the SSA is accurate in comparison with similar numerical simulation approaches (e.g., the systems of ordinary differential equations), it is empirically considered as a computationally expensive algorithm. In this paper, we propose an efficient parallel simulation algorithm for SSA to improve the execution of (coloured) stochastic Petri nets. We illustrate our proposal by applying it to a well-known model in the biological context: reaction diffusion. Furthermore, we evaluate the performance of the parallel algorithm and compare its runtime behaviour with the sequential one.

Optimization becomes a very important methodology appear in scientific life. It can be applied in many different application fields, like telecommunications, data mining, design, combinatorial optimization, power systems and Electronic circuits. Development of electronic circuit is a complex process that needs some simplification that may be difficult to be done using traditional way. In This paper a hybrid rough particle swarm optimization (HRSO) algorithm is proposed for electronic circuit simplification. The (HRSO) is applied to simplify circuit by reducing the components of circuit to try to find optimal value of circuit components. General Terms Soft computing, data mining.

The rapid spread of using Cloud services has become a source of attraction for many business owners due to its multiple potential benefits of flexibility and efficiency with variant pricing plans that can be suited for all types of users. However, there are some vulnerabilities that have important and influential impact on the Cloud Computing efficiency, specially its ability to achieve availability and security for its customers. In this paper we try to find a way for assessing some of public Cloud frameworks for the E-learning application from two points of view; the availability degree of the Cloud framework in both service's speed and cost, and the vulnerabilities and threats that may infect these frameworks. The paper also presents a developed risk assessment equation in attempting to measure the risk factor for each Cloud platform by measuring the impact of the probable threats on the existing vulnerabilities in every platform. The risk rate measurement may help the user in selecting the most proper framework to deploy his application.

The increasing volume of unsolicited bulk e-mail (also known as spam) has generated a need for reliable anti-spam filters. Machine learning techniques now days used to automatically filter the spam e-mail in a very successful rate. In this paper we review some of the most popular machine learning methods (Bayesian classification, k-NN, ANNs, SVMs, Artificial immune system and Rough sets) and of their applicability to the problem of spam Email classification. Descriptions of the algorithms are presented, and the comparison of their performance on the SpamAssassin spam corpus is presented.

The Service Level Agreement (SLA) becomes an important issue especially over the Cloud Computing and online services that based on the ‘pay-as-you-use’ fashion. Establishing the Service level agreements (SLAs), which can be defined as a negotiation between the service provider and the user, is needed for many types of current applications as the E-Learning systems. The work in this paper presents an idea of optimizing the SLA parameters to serve any E-Learning system over the Cloud Computing platform, with defining the negotiation process, the suitable frame work, and the sequence diagram to accommodate the E-Learning systems.

Optimal utilization of resources is the core of the provisioning process in the cloud computing. Sometimes the local resources of a data center are not adequate to satisfy the users’ requirements. So, the providers need to create several data centers at different geographical area around the world and spread the users’ applications on these resources to satisfy both service providers and customers QoS requirements. By considering the expansion of the resources and applications, the transmission cost and time have to be concerned as significant factors in the allocation process. According to the work of our previous paper, a Resource Provision Optimal Algorithm (RPOA) based on Particle Swarm Optimization (PSO) has been introduced to find the near optimal resource utilization with considering the customer budget and suitable for deadline time. This paper is considered an enhancement to RPOA algorithm to find the near optimal resource utilization with considering the data transfer time and cost, in addition to the customer budget and deadline time, in the performance measurement.

As of late, Feature extraction in email classification assumes a vital part. Many Feature extraction algorithms need more effort in term of accuracy. In order to improve the classifier accuracy and for faster classification, the hybrid algorithm is proposed. This hybrid algorithm combines the Genetics Rough set with blind source separation approach (BSS-GRF). The main aim of proposing this hybrid algorithm is to improve the classifier accuracy for classifying incoming e-mails.

The guarantee of delivering the service from the service provider is a vital requirement for the Cloud’s users. This guarantee can be achieved through the Service Level Agreement, which is a contract between the user and the provider. In this paper, an SLA document is defined for an E-learning system called “go learn cloud system – glcs” to clarify the rights, terms and conditions for the users and providers. The SLAin this system has two types; a document between the Cloud service provider and the coordinator of the E-learning system, and a document between the students and the instructor from a side and thecoordinator from the other side. On the other hand, this paper presents how the coordinator of the E-Learning systemscan take hisdecision for choosing the suitable Cloud Computing platformthat can serve the system’s user with minimum cost.

The integration of information and communication technologies in education according to the global trend occupied a great interest in the Arab world through E-Learning techniques and put it into the form of services within Services Oriented Architecture Technique (SOA), and mixing its input and outputs within the components of the Education Business Intelligence (EBI) and enhance it to simulate reality by educational virtual worlds.This paper presents a creative environment derived from both virtual and personal learning environments based on cloud computing which contains variety of tools and techniques to enhance the educational process. The proposed environment focuses on designing and monitoring educational environment based on reusing the existing web tools, techniques, and services to provide Browser-based-Application. I. INTRODUCTION There is much debate among those interested in education and E-Learning in the Arab world about the advantages and disadvantages of this system of education, and most Arabic academic conference papers presented concerned completely with the same topic which is the advantages and disadvantages and the comparison with the traditional educational systems, so in this paper we will move from E-Learning evaluation to fly in the space of implementation.