International Journal of Advanced Networking and Applications

4874-4883

Authors: Erdal Özbay*, Feyza Altunbey Özbay

Melanoma is the most common type of skin cancer due to a genetic predisposition. In recent years, it has been determined that the number of different types of skin cancer has increased worldwide and caused a large number of deaths. Some skin cancers, such as melanoma and its derivatives, can be prevented, but early and accurate diagnosis is very important for treatment. Image processing techniques in medical applications are frequently used in the diagnosis, follow-up, and treatment processes of skin cancer. However, manual control of medical images is laborious and time-consuming and is vulnerable to expert errors in the interpretation of images. Developing a safe and autonomous classification system for medical applications is a fundamental need. In this study, a CNN-based deep learning framework has been developed in which the HAM10000 dataset, a dermatoscopic clinical skin image collection, has been classified for skin cancer detection. Classification preprocessing using contrast limited adaptive histogram equalization is demonstrated by the accuracy results that improve the recognition of subtle features of class labels. A 45-layer model is proposed for classification. With this developed model, an accuracy rate of 99.69% has been achieved. The results show that the proposed model achieves high accuracies and F-measures with low false-negative compared to known classifiers. This CNN model showed the best two-level performance classifying melanoma and benign cases as nevi and non-nevi. It has emphasized that skin cancer can be detected early with the proposed model and can contribute to the execution of the treatment process.

4891-4895

Authors: Raja Marappan

Graph coloring problem, a combinatorial optimization problem is being widely applied in solving the channels allocation in wireless networks. This paper exhibits a new evolutionary genetic multi-objective strategy that uses the combined single and multi-parent conflict-gene crossover and combined single and multi-parent conflict-gene mutation operators with clique partitioning to solve the graph coloring and channel allocation problems. The proposed operators minimize problem search space by reducing the expected number of genetic generations. A general fitness function is defined on finding the total conflicting edges in the graph for the initial and particularly the successive generations of individuals in the population. The outcomes of this proposed method are better than the well-known methods and are compared with some of the benchmark graph coloring and channel allocation problems. The devised method of clique partitioning with genetic operators also enhances the successful runs.

4915-4918

Authors: Ravi Shankar Pandey

Service discovery infrastructure provides automatic integration of web services to meet new business objectives. It has three actors service provider, service broker and service requestor. The broker provides the requested service from the registry. The searching of services is very difficult process. The services should be compatible in terms of functional aspect as well as non functional features. Several research efforts have been made to discover the correct services which are going to integrate. Service requestor may use same set of services many times or different requestor may need same set of service. In this paper, I have proposed a model to support service discovery using the current history of set of services involve in execution of a business objective. The context service graph stores the successful execution of set of web services and stores the QoS at start node for comparing the QoS requirement with the new service requestor QoS. The context service graph is present at the broker side along with task details.

4925-4930

Authors: Amira Hassan Abed

This article centers around the exploration related to the e-learning in the smart cities. The recent innovation, for example, Internet of Things (IoT) is quickly grown in the computerized life. Formation of the intelligent urban communities is developing with the idea of the IoT in the same time. E-residents as the fundamental component play an imperative part in building the keen urban communities. It is undeniable that another type of the resident in the smart cities can assume a fundamental part in case he/she gets satisfactory e-learning. In the computerized life, the IoT grounds in the intelligent urban communities are focused on the enhancement of the e-Leaning part by utilizing advanced communications and methods. Our work here centers around the requirement of embracing IoT techniques in campuses of intelligent cities , as well as supporting the theoretical analysis about the anticipated benefits of the smart learning and its application in the brilliant communities in a definite discussion.

4939-4946

Authors: Amira Hassan Abed, Essam M Shaaban

Building up a breast cancer screening platform is vital to encourage early "Breast cancer" detection and treatment. Proposing a screening system utilizing clinical imaging methodology that doesn't cause body tissue harm (non-obtrusive) and doesn't include actual touch is a major challenge. Thermography, a "non-intrusive" and "non-contact" malignancy screening strategy, can recognize tumors at the beginning phase significantly under determined conditions by noticing temperature circulation in the two bosoms. The thermograms can be deciphered utilizing Deep learning models, for example, "convolutional neural networks (CNN)". CNNs can naturally group bosom thermograms into classifications, for example, ordinary and up-normal. In this work, we intend to cover the most significant studies identified with the usage of deep neural networks for bosom thermogram classification. As we accept that, an overview of breast thermogram possibilities shows that the early manifestations of bosom malignant can be seen by recognizing the asymmetrical warm dispersions between the bosoms. The asymmetrical warm appropriation on bosom thermograms can be assessed utilizing a computeraided platform that depended on deep learning models.