Volume 7 Issue 12
Author's Name: Sivaraj K, Rafik Shafi
Abstract—It is possible to identify organizational characteristics that affect project effectiveness. These components can be recommended for improvement in efforts to build highways. Data from highway construction organizations are acquired via a questionnaire survey to meet the study's goals. The primary elements affecting organizational performance are first noted. The criteria include size, duration, resource availability, complexity, internal and external interdependencies, customer base, uncertainty, and importance. Then, the factors that fall within these primary categories are identified. The efficacy of the organization is a result of all 10 components, some of which are essential to its operation. Because of this, this project might be helpful in enhancing the qualities of organizations that can be used in the future for significant construction projects.
Author's Name: Sohan R. Anup Swarup, Changale
Abstract—Classifying fruits from an image is a remarkable research area when we talk about computer vision. In this regard, this paper proposes a system where fruit detection and fruits classification are performed. In India, fruit yield is turned down because of the post conceding of many spots(disease) on the fruits by the farmers at the end. Agronomists struggle a lot for economical loss across the world. We can say that fruits diseases are the main cause of agricultural loss. To improve their productivity, it is important to know the health status of fruits to help the farmers. This motivates us to design and develop a model to help farmers detect the diseases in the early stage itself. The idea behind this work is to develop automatic models that recognize the disease's level and grade them accordingly. Convolutional neural networks opted for the classification, again retrained using the transfer learning technique. The system is developed in the Tensor flow platform. For the proposed work three to four types of fruits have been considered. The model shows 98.17% training accuracy, whereas, the deep learning-based fruits classification test model shows 99.99% accuracy.
Author's Name: PARASITA , PUNITA MISHRA , SHIVATAJ KUMAR
Abstract—Automatic gender detection through facial features has become a critical component in the new domain of computer human observation. Automatic gender detection has numerous applications in the area of recommender systems, focused advertising, surveillance, biometric and other security systems nowadays. A gender detection system can be implemented in the gender specific areas like woman compartments, gender detection can be used in temples and customized advertisement and in robotics field also. By analyzing the face we can get a lot of information such as, emotions, expressions, skin color, beard, mustache, etc. can be extracted. In recent literature, detection of gender by using facial features is done by many methods such as Gabor wavelets, artificial neural networks and support vector machine, Adaboost, KNN, CNN. This work presents the comparative study of various approaches that are used in gender detection system and their accuracy have been compared
Author's Name: Manmathan , Deepa Ranji , Gajal Mulati
Abstract—The pervasive and wide applications like scene understanding, video observation, advanced mechanics, and self-driving frameworks set off tremendous exploration in the area of computer vision in the latest ten years. Being the center of this multitude of uses, visual acknowledgment frameworks which incorporates picture order, confinement and location have accomplished extraordinary exploration energy. Because of critical improvement in neural organizations particularly profound learning, these visual acknowledgment frameworks have accomplished amazing execution. Object identification is one of these areas seeing incredible accomplishment in computer vision. This paper demystifies the job of profound learning strategies in view of convolutional neural organization for object recognition. Profound learning structures and administrations accessible for object identification are additionally articulated. Profound learning strategies for cutting edge object location frameworks are evaluated in this paper
Author's Name: Manisa Gupta , Kuchi Garg
Abstract— This paper Many applications in computer vision they need precise and efficient detection systems. This demand coincides with the rise of the application of deep learning techniques in almost all areas of machine learning and artificial vision. This work presents a study that encompasses different detection systems based on deep learning, providing a unified comparison between different frameworks in order to carry out a technical comparison of the performance measures of the studied methods Many applications in computer vision they need precise and efficient detection systems. This demand coincides with the rise of the application of deep learning techniques in almost all areas of machine learning and artificial vision. This work presents a study that encompasses different detection systems based on deep learning, providing a unified comparison between different frameworks in order to carry out a technical comparison of the performance measures of the studied methods
Author's Name: Anuj Sharma , Naresh Jadav
Abstract— Health-care organisations may foresee patterns in a patient's medical condition and behaviour by using data mining, which entails examining several options and establishing connections between apparently unrelated bits of information. The volume and variety of raw data generated by healthcareinstitutions make it difficult to make sense of everything. Data must be collected and stored in an organised manner, as well as integrated, in order to develop a unified medical information system. In health, data mining permits the examination of a diverse set ofdata models that are unavailable or obscured by conventional analytical techniques. The objective of this research is to take a diabetic health dataset and analyse it using machine learning techniques to increase diabetesprediction accuracy
Author's Name: Mahi Sewatia , Deepa Rani
Abstract— This paper Many applications in computer vision they need precise and efficient detection systems. This demand coincides with the rise of the application of deep learning techniques in almost all areas of machine learning and artificial vision. This work presents a study that encompasses different detection systems based on deep learning, providing a unified comparison between different frameworks in order to carry out a technical comparison of the performance measures of the studied methods Many applications in computer vision they need precise and efficient detection systems. This demand coincides with the rise of the application of deep learning techniques in almost all areas of machine learning and artificial vision. This work presents a study that encompasses different detection systems based on deep learning, providing a unified comparison between different frameworks in order to carry out a technical comparison of the performance measures of the studied methods
Author's Name: Asharad Sharma , Chandraprakash Yadav , Garima Nazim
Abstract— This paper presents In recent days statistical reports focus those 8 to 10 conditions in kids whenever analysed early can forestall youth visual impairment which incorporates youth Strabismus (Squint Eye), Amblyopia (Lazy Eye) and so on, metropolitan regions have 1 ophthalmologist for 10,000 individuals yet in country zones, it is 1 for every 2,50,000 individuals. This paper provides a systematic study based mainly on strabismus detection surveys using Machine Learning techniques. Among all Machine Learning Algorithms Convolution Neural Network - CNN produces more accuracy of 95.83%. than Support vector machine -SVM.
Author's Name: Vaidehi, Uma Macchani
Abstract— The Cloud Computing is enabling innovative and on demand services by allowing pay per use, location independency and device independency. In process of migration VM moves one physical machine to another. In live migration, the VMs are migrated without stopping their working. In offline migration, process is stopped till the VM can continue on target machine. In this we first present a live migration performance strategy. Live Task of migration and the needed properties of VM for monitoring the resources and optimal the fitness function evaluation. We Will reduce the operation cost, down time and also increases the resource utilization than migrate VM one server to another server base checking of prediction capacity. To solve the problem of the overload of virtual machine, virtual machine migration techniques used which maintain the load balance on the Physical Machine which is undergo unnecessary problems caused during the time of overload and also optimize the resource utilization and total down time
Author's Name: Sachin kumar, Subender Madam
Abstract— proposed project is dedicated towards the designing and developing of an indoor navigation system with the most optimum characteristics. Global Positioning system is suitable for indoor navigation system within contact of building blueprint. The indoor navigation system is purely based on the application of mobile network found abundantly in smart phones. The main goal of proposed work is to minimize the cost and maximize the end user benefits with more efficiency. The system has been provided with two main functions which are to provide localisation and navigation services. Navigation entails the continuous tracking of the user’s position and surroundings for the purpose of dynamic planning and following a route to the user’s intended destination. Indoor navigation system has wide application in the industry and home automation field. The application has been developed for both iPhone and Android in order to compare the results and see if one operating system is better than the other for the purpose of this application. The proposed navigation system for smartphone is capable of guiding users accurately to their destinations in an unfamiliar indoor environment, without requiring any expensive alterations to the infrastructure or any prior knowledge of the site’s layout.
Author's Name: R Aarav Krishna, Auyan Yusuf
Abstract— In current engineering practice the design methods for earth retaining walls under seismic conditions are mostly empirical. Dynamic earth pressures are calculated assuming prescribed seismic coefficient acting in the horizontal and vertical directions using time history analysis Structural dynamic deals with method to determine the stresses and displacement of structure subjected to dynamic loads .the dimension of structure are finite. It is thus rather straight forward to determine dynamic model with finite no of degree of freedom. The corresponding dynamic equation of motions of the discretized structure is then formulated, and highly developed methods for solving them are radially available) In this study nonlinear analysis of retaining wall is studied including soil structure interaction for various type of walls for silty soil, clay soil and sandy soil . The data collected for time history analysis is Koyana,Bhuj, Kobe,Uttrakashi and El Centro. The software used for analysis is ANSYS in which we can model any type of material for soil structure interaction upon this study.
Author's Name: Purav Karan,H.H. Duggu, S. Mithran
Abstract— The aim of the present study is to compare the behavior of multi-storey building having flat slab with or without shear walls and to analyze the effect of building height on the performance under earthquake forces. Also effect of with or without shear wall for flat slab building on seismic behavior with varying thickness and varying position of shear wall are studied. In this work, the effects of varying seismic zones on these buildings are also carried out. For that, G+9 and G+19 Storey models, each of plan size 20X20m are selected. For stabilization of the variable parameters, shear wall are provided at corners, center and along the periphery. To study the effect of varying thickness and different location of shear wall on flat slab multi-storey building, static analysis (Equivalent Static Analysis) in software STAAD Pro is carried out for zone IV and V. The seismic parametric studies comprise of lateral displacement, storey drift, drift reduction factor and contribution factor.
Author's Name: Sridaan Nayak, Ishaan Singh, Omkar Atharva
Abstract— — Over the years, several researchers have been interested in the natural intestinal flora. In the 1920's, James Reyniers' pioneering work created the first sterile guinea pig. Comparing the physiology of sterile bacteria and traditional farm animals offers useful knowledge on how host biology affects the gut flora. We now know that the intestinal flora can control the immune system, proliferation of epithelial cells, angiogenesis of the intestines, development of hormones, energy absorption and conduct. Moreover, recent studies indicate that obesity is associated with changes in the intestinal flora and the use of sterile mice has shown the direct involvement of the microbial flora in disease development.
Author's Name: Yahya Pransh, Shreyansh Kushaan, Sri Pranav
Abstract— — Irrigation in India includes a network of major and minor canals from Indian rivers, groundwater well based systems, tanks, and other rainwater harvesting projects for agricultural activities. Of these groundwater system is the largest In 2013-14, only about 47.7% of total agricultural land in India was reliably irrigated. The largest canal in India is Indira Gandhi Canal, which is about 650 km long. About 2/3rd cultivated land in India is dependent on monsoons. Irrigation in India helps improve food security, reduce dependence on monsoons, improve agricultural productivity and create rural job opportunities. Dams used for irrigation projects help produce electricity and transport facilities, as well as provide drinking water supplies to a growing population, control floods and prevent droughts. India's irrigation covered crop area was about 22.6 million hectares in 1951, and it increased to a potential of 90 mha at the end of 1995, inclusive of canals and groundwater wells. However, the potential irrigation relies of reliable supply of electricity for water pumps and maintenance, and the net irrigated land has been considerably short. According to 2001/2002 Agriculture census, only 58.1 million hectares of land was actually irrigated in India. The total arable land in India is 160 million hectares (395 million acres). According to the World Bank, only about 35% of total agricultural land in India was reliably irrigated in 2010. The ultimate sustainable irrigation potential of India has been estimated in as1991 United Nations' FAO report to be 139.5 million hectares, comprising 58.5 mha from major and medium river-fed irrigation canal schemes, 15 mha from minor irrigation canal schemes, and 66 mha from groundwater well fed irrigation. India's irrigation is mostly groundwater well based. At 39 million hectares (67% of its total irrigation), India has the world's largest groundwater well equipped irrigation system (China with 19 mha is second, USA with 17 mha is third). India has spent 16,590 crore on irrigation development between 1950 and 1985. Between 2000-2005 and 2005-2010, India proposed to invest a sum of 1,03,315crore and 2,10,326 crore on irrigation and flood control in India
Author's Name: Sampath Shakti, Sharma Sikandar
Abstract— In the fast growing internet world, web content increase day by day. This demands the knowledge searching and reasoning in this big data. The knowledge is represented in the semantic web using web ontology languages. Existing methods take long time to derive inferences and also it performs full reasoning when new data stream arrives. In this paper an Incremental Ontology Inference (IOI) Method for handling large number of triples (subject, predicate, and object) is proposed. In IOI, the triples for each type are collected and a forest like data structure is built and then performs reasoning. The storage requirement is also reduced by merging the triple reasoned from other triple into a set of triples with the same values. Hence, it provides fast traversal of triples in the tree and retrieves the query results efficiently. MapReduce paradigm is used to implement the proposed approach. The results for user query are reasoned and retrieved effectively.
