MIT-World Peace University (MIT-WPU) is a top educational institute located in Pune, India. It is recognized within&nbsp;UGC under Government of Maharashtra Act No. XXXV 2017.&nbsp;

It is part of the MIT group of institutions (established in 1983) which covers 10+ campuses across India and has more than 50,000+ students enrolled across various disciplines at a given point of time. The institute houses 15 schools ranging from Engineering &amp; Technology, Design, Pharmacy, Commerce, Governance among others. It offers undergraduate, postgraduate and PhD programs among different disciplines.&nbsp;

&nbsp;

MIT World Peace University

School of Computer Engineering & Technology

Cloud computing that is based on Infrastructure as a Services model, allows the system administrators to identify resources or services to deploy their applications. Selection of the most appropriate provider for a particular application is a difficult task as there are huge numbers of services offered by considerable number of providers that are not directly comparable and have similar functionality. Additionally these systems must comply with non-functional requirements as they evolve. Certain traditional techniques are employed to verify the non-functional properties of critical systems at design time. In this paper, a new framework is suggested that supports the process of deploying applications in cloud providers, using a comparison of high level requirements predominantly non-functional requirements (NFR). Experiments have already been conducted with conclusive results wherein matching of user constraints during service selection process is done. The experiment uses Aspect Oriented Paradigm (AOP), XML and graph transformation. The relevance of the framework is illustrated using the Remote Patient Monitoring (RPM) scenario. The results demonstrate that the model considerably improves the process by identifying conflicting NFR. Based on this work done, the next logical step is to extend it to Cloud Computing. © 2016 IEEE.

Engineering and Technology

Proceedings - 2016 5th IEEE International Conference on Cloud Networking, CloudNet 2016

Validation of non-functional requirements in cloud based systems (short paper)

Content-based image retrieval system nowadays use color histogram as a common color descriptor. We consider color as one of the important features during image representation process. Different transformations such as changing scale of image, rotating an image and translations of image to other forms does not make any alterations to the color content of image. If we need to focus on differentiation or similarity between two images we usually deal with various color features of image. To extract color features of image we consider on color space, color reduction, color feature extraction process. In image retrieval applications, user specifies desired image as query image and wants to search for the most similar image in database of his interest. Application then identifies similar relevant images from database based on different color features of database images and query image. To achieve this we compute color features of database images and those for query image. We use local color features of different regions and combine them to represent color histogram as a color feature. These color features are compared using Euclidean distance as a metric to define similarity between the query image and the database images. For calculations of local color histogram we divide image into different blocks of size 8 × 8 as fixed, so that for each block of image spatial color feature histogram of image is obtained. Our experimental work shows that local hybrid color histogram produced more accurate image retrieval results than global color moments color histogram.

World University root node

Advances in Intelligent Systems and Computing

Local and global color histogram feature for color content-based image retrieval system

Due to robustness feature color of object has been proved as low level unique feature. If we change viewpoint looking at a color we note change in its appearance. Color image processing requires dealing with each and every pixel of an image. In this paper spatial color distribution features are implemented. We focus on its calculation and comparison of retrieval efficiency using histogram quadratic distance measure for HSV color Space. Our experiments show hybrid histograms produced better results.

Proceedings of the 2015 International Conference on Green Computing and Internet of Things, ICGCIoT 2015

Comparison of spatial color histograms using quadratic distance measure

International Journal of Computer Science and Mobile Computing

Stock market forecasting techniques: literature survey

Nowadays there is a huge increase in the use of sensors in various applications such as remote monitoring of environment, automobiles, disaster prone zones, home control, and military applications. The capabilities of Wireless Sensor Network (WSN) can be extended using self-organization to change their behavior dynamically and achieve network wide characteristics. Clustering techniques show the self-organized behavior of WSNs. Sensor nodes are grouped into disjoint, nonoverlapping subsets called clusters. Cluster Heads (CHs) collect data from the sensor nodes present in the cluster and forward it to the neighboring nodes using shortest path distance calculation and finally to the Base Station (BS). In the proposed clustering technique, network area is divided into regions. Cluster Head (CH) is elected by using the highest residual energy and the node degree. Cluster communication is at the most two-hop which results in less number of messages from member nodes to BS. Within a cluster, data is sent to CH by member nodes. After some time CH’s energy level goes below threshold value, then new CH is elected and clusters are reformed within each region. Our proposed clustering technique can be used in military applications to detect and gain information about enemy movements. Results are obtained by varying the number of nodes at different transmission ranges. The simulation results show that the proposed clustering algorithm reduces energy consumption, prolongs network lifetime, and achieves scalability. © Springer Science+Business Media Singapore 2016.

Region-based clustering approach for energy efficient wireless sensor networks

For better transportation system it is needed to make our traffic signal well instructed with the help of Vehicle-To-Infrastructure (V2I) communication for transferring the information about vehicles to the central system. This paper presents the centralized approach in three phases. First phase gives the complete information of Central System (CS), second phase shows the working of Area Controller Systems (ACS) and Area Border Controller Systems (ABCS) and phase three shows the Onsite Unit (OU). © 2016 ACM.

Proceedings of the Second International Conference on Information and Communication Technology for Competitive Strategies

Centralized Approach towards Intelligent Traffic Signal Control

For privacy protection, many architectures such as trusted third party, cryptography, client side privacy etc. have been proposed. These methods are neither fully support privacy nor efficient. We have proposed limited broadcast based application and one dynamic protocol for cloaking. We have simulated user broadcast model, to achieve location privacy where users are mobile and points of interests (POI) are stationary. Second experiment is a POI broadcast model to achieve location privacy. We have also presented dynamic cloaking protocol that considers speed, direction & location of mobile user for cloaking. We have experimentally recorded the results related to performance, energy and privacy strength. It shows that our system has more social and human value along with commercial attraction. © 2016 The Authors.

Fulltext

Procedia Computer Science

Controlled Broadcast Protocol for Location Privacy in Mobile Applications

Theoretical Computer Science

Today ovarian cancer is second most perilous cause of cancer deaths in women after breast cancer. In this work, we have developed system which acquires ultrasound images and using image processing and machine learning algorithms accurately classify benign and malignant tumors in ovarian cancer. This technique denoise image using wavelet transform, grey level texture features extracted using GLCM(grey level co-occurrence algorithm), extracted features will be trained through SVM(Support vector machine) and selected non-redundant features selected through Relief-F will be further train and test through SVM for output. Proposed technique was validated by 60 malignant and 60 benign images of patients. On evaluating classifier for 14-texture descriptors give 74% and relief-F gives 82% accuracy. After selecting 6 features from 14 features it will give accuracy 86% and relief-F gives 92% accuracy. Thus, the features are significant for result and preliminary results depict that the proposed technique can be reliable for ovarian tumor classification as this system is fully automated, advantageous and cost-effective too. © 2015 IEEE.

Proceedings of 2015 IEEE International Conference on Research in Computational Intelligence and Communication Networks, ICRCICN 2015

Identification of ovarian mass through ultrasound images using machine learning techniques

Random Forest is an ensemble, supervised machine learning algorithm. An ensemble generates many classifiers and combines their results by majority voting. Random forest uses decision tree as base classifier. In decision tree induction, an attribute split/evaluation measure is used to decide the best split at each node of the decision tree. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation among them. The work presented in this paper is related to attribute split measures and is a two step process: first theoretical study of the five selected split measures is done and a comparison matrix is generated to understand pros and cons of each measure. These theoretical results are verified by performing empirical analysis. For empirical analysis, random forest is generated using each of the five selected split measures, chosen one at a time. i.e. random forest using information gain, random forest using gain ratio, etc. The next step is, based on this theoretical and empirical analysis, a new approach of hybrid decision tree model for random forest classifier is proposed. In this model, individual decision tree in Random Forest is generated using different split measures. This model is augmented by weighted voting based on the strength of individual tree. The new approach has shown notable increase in the accuracy of random forest.