This paper is based upon the recognition of either characters or digits by using Neocognitron algorithm, which can be used for detecting multiple patterns. A large pattern can be divided into multiple parts until it becomes a single cell for each and every block for its representation. Neural Network can be implemented with the concept of fuzzy logic which deals with ambiguous problems. The integration of Neural Network and Fuzzy Logic will generate such type of intelligent systems for mankind in optimum real life problems. It also describes various types of architects that can be constructed with neural network. In Neocognitron approach, a type of hierarchal network where two types of cells can be used like simple and complex cells, which can be trained by using different neural network algorithm approaches. It also describes various types of applications that can be used in real life through Neocognitron. The objective of the paper is to describe various architectures of Neural Network approach, so that it can be implemented with different algorithms for better outcomes in real life complex problems.
Attribute selection is the procedure of selecting a subset of important attributes for utilization in model development. The central supposition when utilizing as attribute selection method is that the information contains numerous redundant or irrelevant attributes. Repetitive attributes are those which give no more data than the right now chosen attributes, and irrelevant attributes give no helpful data in any setting. Attribute selection is a process in which subset of important attribute is selected which produce good results. Attribute selection algorithm is used for that purpose which achieve efficiency, i.e. less time and correctness of subset. Existing system proposed clustering based attribute selection algorithm based on efficiency and effectiveness criteria. In this algorithm, attributes are first separated into different clusters using graph theoretic clustering method and then those attributes are selected from each clusters, which is most related to target class. Because of large attributes minimal in graph many nodes are generated and in such situations working of prims algorithm is better. In this paper, the system uses Kruskals algorithm instead of Prims algorithm for better efficiency and accuracy. The Kruskals algorithm perform sorting according to the weight and starts from the smallest one which will take less time to iterate. This is the only method which uses sorting technique which will increase the efficiency.
Cloud computing is becoming a widely used technology nowadays. It has changed the way the applications and data area accessed. Rather than using personal computer or local server, cloud computing uses network of remote servers which is on the internet to manage, protect, and store. E-learning based on clouds is possible with a certain payment. It requires many hardware and softwares. By the use of cloud, the process of accessing data has changed a lot. E-learning service consumers like colleges and universities have to pay for the quality resources, utilized for actual duration of time. It is in the same way to pay the electricity bills and water bills according to consumption. This paper presents what cloud computing is, the issues concerning the implementation of cloud-based e-learning, and suggests architecture for elearning which can be used by educational institutes as well as by small and medium scale organizations.
In this paper, a study has been done on improved performance of reactive ad-hoc routing protocol, i.e Dynamic Source routing protocol in context of increasing number of failure mobile nodes in the network. And simulations are evaluated using Network Simulation (NetSim). Mobile Ad Hoc Network consists of mobile sites that are not having any predefined structure. Because of varying infrastructure these nodes are attacked and may get failure also. In present study, reactive routing protocol, i.e. Distance Vector Routing (DSR) is evaluated on the basis of throughput, packets transmitted, packets errored, and packets collided by increasing number of failed nodes in the network.
Slant correction is generally performed to normalise handwritten characters to improve the results of recognition. But, natural scene images containing printed text suffer from both skew and slant deformation due to 3D tilt and projective transformation. In this study, a slant correction technique for natural scene images containing Gurmukhi text words has been proposed.The skew normalized Gurmukhi words are firstly transformed to salient image and Hough transform is then applied onto edge image obtained through Sobel operator in vertical direction. Of all the lines identified by Hough transform, the angle of longest line with vertical axis is inferred as the slant angle of Gurmukhi word. Horizontal affine shear transformation with previously determined slant angle is carried out to correct slant in word image. Experimental results show that the method is very effective for slant correction of Gurmukhi words from scene images. The method works equally well for other headline based Indian scripts like Devanagari as well without any modification.
The utility of Hospital Information System (HIS) is vast, although one cannot ignore its challenges which prevent the electronic healthcare system from being used properly. Challenges in privacy and security of HIS needs to be studied and understood properly by any healthcare organization and should be resolved to get optimum benefit. One of the main objectives of this paper is to review, explore and analyze the current state of hospital information system's privacy and security of patient's electronic health records. It also focuses on security at the level policy of healthcare organization so that electronic patient record can be protected and secured. In healthcare organization security risks and financial consequences are increasing day-by-day. The vulnerable and breached electronic patient data revealed the fact that, the privacy and security of electronic patient records in health information exchanges is an imperative of any healthcare organization. For any patient care organization persons who are involved in the IT and administrative management should seriously think the issues of privacy and security of patient health record and also proper health information exchange in a secured manner.
Cloud Computing (CC) is a buzzword, which came into existence after decades of research using existing technologies like parallel computing, grid computing, peer to peer technology, distributing computing, and virtualization. Now-adays, the most approved applications are based on internet services with a large number of users. Therefore, as the size of cloud scales up, cloud computing service providers make it necessary to handle massive requests. Load balancing is one of the main challenges in cloud computing which distributes the dynamic workload across multiple nodes to ensure that no single resource is either overburdened or under used (idle). Therefore, several load balancing algorithms have been designed, which reduces the response time of different tasks and also improve the resource utilization. So, a Dynamic threshold based load balancing algorithm has been developed in this paper. This algorithm distributes the load among different virtual machine as well as improves the response time. Thus the proposed algorithm improves the precision of load prediction. Additionally, Cloud Analyst tool is used to simulate this work and then compare with AMLB (Active Monitoring) and Throttled load balancing algorithms. Results demonstrate that the performance of proposed algorithm is better than AMLB and Throttled load balancing algorithm in terms of load and response time.
