At present days, DR becomes a more common disease affecting the eyes because of drastic rise in the glucose level of blood. Almost half of the people under the age of 70's get severely affected due to diabetes. The earlier recognition and proper medication results to loss of vision in several DR patients. When the warning signs are identified, the severity level of the disease has to be validated to take decisions regarding the proper treatment. The current research focuses on the concept of classifying the images of DR fundus based on the severity level using a deep learning model. This paper proposes a deep learning based automated detection and classification model for fundus diabetic retinopathy (DR) images. The proposed method involves several processes namely preprocessing, segmentation and classification. Initially, preprocessing stage is carried out to get rid of the unnecessary noise exist in the edges. Next, histogram based segmentation takes place to extract the useful regions from the image. Then, synergic deep learning (SDL) model is applied to classify DR fundus images to various severity levels. The justification of the presented SDL model is carried out on Messidor DR dataset. The experimentation indicated that the presented SDL model offers better classification over the existing models.

No abstract

This paper compares various selection techniques used in Genetic Algorithm. Genetic algorithms are optimization search algorithms that maximize or minimizes given functions. Indentifying the appropriate selection technique is a critical step in genetic algorithm. The process of selection plays an important role in resolving premature convergence because it occurs due to lack of diversity in the population. Therefore selection of population in each generation is very important. In this study, we have reported the significant work conducted on various selection techniques and the comparison of selection techniques.

Wireless Sensor Networks (WSNs) have left an indelible mark on the lives of all by aiding in various sectors such as agriculture, education, manufacturing, monitoring of the environment, etc. Nevertheless, because of the wireless existence, the sensor node batteries cannot be replaced when deployed in a remote or unattended area. Several researches are therefore documented to extend the node's survival time. While cluster-based routing has contributed significantly to address this issue, there is still room for improvement in the choice of the cluster head (CH) by integrating critical parameters. Furthermore, primarily the focus had been on either the selection of CH or the data transmission among the nodes. The meta-heuristic methods are the promising approach to acquire the optimal network performance. In this paper, the 'CH selection' and 'sink mobility-based data transmission', both are optimized through a hybrid approach that consider the genetic algorithm (GA) and particle swarm optimization (PSO) algorithm respectively for each task. The robust behavior of GA helps in the optimized the CH selection, whereas, PSO helps in finding the optimized route for sink mobility. It is observed through the simulation analysis and results statistics that the proposed GAPSO-H (GA and PSO based hybrid) method outperform the state-of-art algorithms at various levels of performance metrics.

No abstract