Journal of Analysis and Computation
Journal of Analysis and Computation
POTATO DISEASE DETECTION USING MACHINE
LEARNING
Grusha Dadiyal¹, Harsha¹, Khushi Bhushan¹, Abhilasha Borkar²
¹Student, Department of Computer Engineering, Cummins College of Engineering for Women, Nagpur, India
²Assistant Professor, Department of Computer Engineering, Cummins College of Engineering for Women, Nagpur, India
Keywords – Convolutional Neural Network, Image classification, K-means, Support Vector Machine (SVM).
ABSTRACT
Potatoes are a major crop in India, grown year-round due to their economic value and widespread consumption. Their ability to grow in various soil conditions makes them appealing to farmers. However, potato production is hindered by diseases, increasing costs and disrupting farmers’ lives. To address this, we propose an automated software solution using advanced machine learning techniques such as data augmentation and Convolutional Neural Networks (CNN). By analyzing leaf pictures, we can detect and differentiate between early blight (caused by a fungus) and late blight (caused by a specific microorganism). Early detection allows farmers to apply appropriate treatment, minimizing waste and economic loss. We trained our model using 2,152 images of healthy and diseased potato leaves, achieving a 99.25% accuracy rate on the test set. Our goal is to process, train, and model the data efficiently for maximum output.
EFFECT OF VARIATION OF INCLINATION ANGLE OF
DIVERGING PORTION OF COMPOSITE WIND DIFFUSER ON
VIBRATION CHARACTERISTICS
Shinde Sampada Vijay¹, Prof. S. S. Adewar²
¹Student, ²Professor, ME Design, Department of Mechanical Engineering,
1,2Department of Mechanical Engineering, Zeal College of Engineering and Research, Pune-411041,
(India)
Keywords: DAWT, FEA, FFT, analyzer and impact hammer test.
ABSTRACT
Wind energy technology is one of the fastest growing alternative energy technologies. However, conventional turbines commercially available in some countries are designed to operate at relatively high speeds to be appropriately efficient, limiting the use of wind turbines in areas with low wind speeds, such as urban areas. Therefore, a technique to enhance the possibility of wind energy use within the range of low speeds is needed. The techniques of augmenting wind by the concept of Diffuser Augmented Wind Turbine (DAWT) have been used to improve the efficiency of the wind turbines by increasing the wind speed upstream of the turbine. DAWTs are studied in terms of diffuser shape design, sizing of investigation and geometry features which involved diffuser length, diffuser angle, and flange height. Our main objective of this project is to optimize the Diffuser design of DAWT and to determine the maximum velocity obtained inside the diffuser for placing the blades to achieve more power output. Finding the effect of changing diffuser angle & material of diffuser on vibration characteristics. Experimental validation of optimize diffuser model will perform on FFT analyzer and impact hammer test.
EVALUATION OF MECHANICAL PROPERTIES OF
POLYPROPYLENE INCORPORATING WITH MOLYBDENUM
DISULPHIDE
Bangale Abhijit A.¹, Dr. Magade Pramod. B.², Shevale Sanket S.³
1,2Department of Mechanical Engineering, ZCOER, Pune-411046, (India)
³Skills, Employment, Entrepreneurship and Innovation Department, Govt. of Maharashtra (India)
Keywords: Polypropylene (PP), Molybdenum disulphide (MoS2), Injection Moulding, Extrusion Moulding.
ABSTRACT
The findings of the conducted investigation validate the effectiveness of utilizing injection molding and extrusion techniques in producing blends of polypropylene (PP) and Molybdenum disulfide (MoS2) polymer matrix, resulting in the development of innovative composite materials through a favorable melt-processing approach. These two methods are predominantly employed for manufacturing specific components of devices on a large scale. The impact of varying compositions of MoS2 on the performance of Polypropylene material was examined through direct extrusion and injection molding of sample specimens. Tensile, compression, flexural, Izod impact, and micro hardness tests were performed on the samples containing different MoS2 compositions ranging from 0.93 to 7.12 vol.%. It was observed that the Molybdenum disulfide (MoS2) particles were uniformly distributed within the polypropylene (PP), demonstrating strong bonding between the two components. The inclusion of MoS2 led to significant enhancements in performance through reinforcing effects, highly efficient nucleation activity, and favorable bonding characteristics. However, composites with MoS2 content above 4.28 vol.% exhibited a decline in mechanical properties compared to those with lower MoS2 content.
ANALYSIS AND EXPERIMENTAL INVESTIGATION OF
FORCE SUSTAINABILITY OF KENAF-COMPOSITE FIBER
LAMINATES
Ms. Prajakta B. Bhawal¹, Prof. Shraddha S. Adewar², Dr. Pramod.B. Magade³, Prof. Sachin M. Godase4
1PG Student, 2Assistant Professor, Department of Mechanical Engineering, ZCOER, Pune
Keywords: Kenaf Fiber, Carbon Fiber, Reinforecement.
ABSTRACT
This work presents the study of tensile properties of composites made from, Kenaf & carbon fiber composite material. The composite material are prepared using hand lay up techniques for different ply orientation of 0º, 30º, 45º & 60º. Initially design of standard specimen in CATIA software along with analysis in ANSYS software to observe the effect of stress concentration on each layer by layer using ACP tool post module in ANSYS 19.0. This project aim is to development of a new hybrid composite made of kenaf include in carbon epoxy composite so that the extent of utility of the newly developed composite could be established. Properties considered for characterization of composite Tensile strength. This research indicates that tensile strength is mainly dependent on the fiber orientation.
ARTIFICIAL INTELLIGENCE BROWSER USING JAVA SPEECH
Y. Sri Lalitha
Associate Professor, Department of IT, Gokaraju Rangaraju Institute of Engineering and Technology
Keywords: AI, Java speech, internet browser, voice commands.
ABSTRACT
This paper is used for browsing websites on the internet by the disabled people with the help of voice. They can browse sites by speaking the site name which will get printed on the address bar with the help of voice recognizer. The page which gets opened can be read to the person by using voice synthesizer which converts text into speech. The basic problem is that disabled people cannot make use of the internet efficiently as they require. The speech recognizer provides them a platform to work over internet through speech. The browser that is developed is a voice browser which recognizes the speech and loads that particular webpage. The displayed content can be read to the user with the help of voice synthesizer which converts text into speech. Certain voice commands are used for this purpose. This browser also provides commands to navigate through the pages.
SOLID WASTE MANAGEMENT THROUGH AUTOMATIC
GARBAGE MONITORING AND COLLECTING SYSTEM
Sagar Shinde¹, Lalit Kumar Wadhwa²
¹JSPM Narhe Technical Campus, Pune, India
²Dr. D. Y. Patil Institute of Engineering & Technology, Pimpri, Pune, India
Keywords: Garbage Collection, Sensors, ARM Controller, Webcam, GSM.
ABSTRACT
In day-to-day life the solid waste management plays a vital role. It will directly or indirectly effect on health issues of human being and animals. The monitoring of garbage collection and its management is important in each and every city and villages as well. It is need to develop the system that can helps administrators in decision making related to reallocation of paths and containers etc. and handling management issues such as observing performance of contractors and waste generation characteristics of a specific area. It also enhances transparency in the working of the civic administration. The proposed system would be able to monitor the garbage tracking process and manages the comprehensive collection process. It would provide in time garbage tracking & collection through the MATLAB database and also gives advantages such as usage of minimum route, low fuel cost, clean environment and available vehicle. The system provides real time monitoring of garbage container through a webcam-based application & GSM.
REAL TIME DETECTION, CONTROLLING OF RAILWAY
ACCIDENTS & ENERGY CONSUMPTION
Sagar Shinde¹, Lalit Kumar Wadhwa²
¹JSPM Narhe Technical Campus, Pune, India
²Nutan Maharashtra Institute Of Engineering & Technology, Pune
Keywords: WSN, Energy Consumption, Load Cell, Vibration Sensor, Rail Accidents.
ABSTRACT
One of the most important factors that railway operators need to consider when it comes to safety is track breaking. This issue can lead to various accidents. In real time, this problem can be identified and fixed before a train actually approaches the broken track. In order to address this issue, a vibration sensor has been developed that can detect cracks in the railway tracks. This system can also be used to monitor the condition of the tunnel lights. The load cell and switching circuitry of this system can be utilized to switch on and off the lights when the train is entering or leaving the tunnel.
A SMART ELECTRONIC SYSTEM FOR REAL TIME
DETECTION, MONITORING & CONTROLLING ROAD
ACCIDENTS
Sagar Shinde¹, Lalit Kumar Wadhwa²
¹JSPM Narhe Technical Campus, Pune, India
²Nutan Maharashtra Institute Of Engineering & Technology, Pune
Keywords: Alcohol Sensor, Road Accidents, IR Sensor, Heart Rate Monitoring, Fuel Sensor.
ABSTRACT
Road safety is a major concern in today’s world. Accidents happen frequently on highways and these cause a lot of problems. Some of the factors that can cause these accidents include speeding, drunk driving, distracted driving, and fuel leakage. Another factor that can cause these accidents is when another vehicle gets in front of the car that is traveling in the same lane. Cars are known to be the most critical systems in the world due to the way they are used. In addition to being used for various functions, intelligent systems are also used in cars to improve their safety. The current system for controlling accidents is not very effective and it has a spineless solution. If an accident occurs, the location of the wreck is not known. A proposed solution that would allow four-wheelers to monitor their surroundings and prevent bad driving that causes accidents would be equipped with sensors that can detect alcohol and to reduce the menace and removing impairment of bad driving. This system would also function as a car black box. It would be equipped with various sensors that can be used to analyse the cause of the accidents. These include gas sensors, a temperature sensor, and a heart rate monitor.
AN EFFICIENT PARALLEL METHOD FOR MINING FREQUENT CLOSED
SEQUENTIAL PATTERNS
Priyanaka patil¹ and Dinesh D. Patil²
¹PG Scholar Student, ²Associate Professor and Head Department of Computer Science and Engineering, Shri. Sant Gadge Baba College of Engineering Bhusawal, India
Keywords: Candidate Pruning, Frequent Itemset, Utility mining, Data Mining, High utility itemset, Dynamic Bit.
ABSTRACT
Data Mining is the process of mining data from different resources and extract data from the relevant information. It can be defined as the activity that extracts information contained in very large database. That information can be used to increase the revenue or cut costs. Association Rule Mining (ARM) is finding out the frequent itemsets or patterns among the existing items from the given database. High Utility Pattern Mining has become the recent research with respect to data mining. Algorithm is use to incremental database and also the support counts of newly generated frequent itemsets. It is not only includes new itemset into a tree but also remove the infrequent itemset from a utility pattern tree structure. Hence, it provides faster execution, that is reduced time and cost. We propose an effective parallel approach in which we are going to combine the Apriori algorithm with Dynamic Bit factor Method.
REAL TIME MULTISENSOR TECHNIQUE FOR FALSE ALARM
REDUCTION USING IOT
Hetusha Patel¹, Suhani Patel², Prof. Meet Shah³, Dr. Lokesh Sharma4
¹patelhetusha@gmail.com , ²sonapatel2323@gmail.com, ³shahmeetk@yahoo.co.in, 4lksharmain@gmail.com
¹Rollwala Department of Computer Science, Gujarat University, Ahmedabad.
Keywords: Temperature sensor, wireless sensor networks, fuzzy logic, Moisture sensor, Smoke sensor, Flame Sensor, False Alarm reduction, PNN.
ABSTRACT
The present development number of worldwide systems administration enormously affects the participation of shrewd components, of discretionary kind and reason that can be found anyplace and cooperate with each other as indicated by the predefined convention. Besides, these components must be shrewdly coordinated with a specific end goal to help disseminated detecting as well as checking/control of true wonders. That is the reason the Internet of Things (IoT) idea raises like another, promising worldview for Future Internet advancement. Considering that Wireless Sensor Networks (WSNs) are imagined as basic piece of subjective IoTs, and the conceivably immense number of collaborating IoTs that are normally utilized as a part of this present reality wonders checking and administration, the unwavering quality of individual sensor hubs and the general system execution observing and change are certainly testing issues. A standout amongst the most intriguing true marvels that can be observed by WSN is indoor or open air fire. The joining of delicate processing advancements, as fluffy rationale, in sensor hubs must be explored keeping in mind the end goal to pick up the reasonable system execution observing/control and the maximal expansion of segments life cycle. Numerous viewpoints, for example, courses, channel get to, finding, vitality effectiveness, scope, arrange limit, information collection and Quality of Services (QoS) have been investigated broadly. There are two fluffy rationale approaches, with transient attributes, are proposed for observing and deciding certainty of flame keeping in mind the end goal to advance and lessen the quantity of standards that must be checked to settle on the right choices thus we are going contribute our idea Smoke Sensor, Flame Sensor, Moisture Sensor. We accept that this decrease may bring down sensor exercises without applicable effect on nature of activity and diminish the likelihood of false alert contributing the proficiency, power and cost viability of detecting keeping in mind the end goal to get an ongoing check of proposed approaches a model Sensor web hub, in view of Representational State Transfer (RESTful) administrations, is made as a foundation that backings quick basic occasion flagging and remote access to sensor information through the Internet A probabilistic neural system (PNN) is a sustain forward neural system, which is generally utilized as a part of arrangement and example acknowledgment issues And furthermore used to diminish FAR.
ACCIDENTS VEHICAL ANALYZER USING BIG DATA
Shipra Soni¹, Ankur Goyal²
¹Research Scholar Department of Computer Engineering
²HOD Department of Computer Engineering Yagyavalkya Institute of Technology Jaipur, India
Keywords: Data Analysis, Hadoop, Map Reduce, PIG, Heatmap, Android.
ABSTRACT
The aim of the work described in this paper is to obtain a greater understanding of the problem of the road accidents in developing and developed countries like India by different vehicle types. Road accidents are serious problem and one of the crucial areas of research in the world. A Road accident causes injuries, disabilities and even fatalities. The total number of fatal accidents as well as related fatality in the country is increasing over the years. To understand the cause, we need the traffic dataset to analyze it. The traffic accidents dataset are very large as the volume of traffic increasing day by day, it gets difficult to store, process and predict such huge sets of data using traditional software. Hadoop is one such framework which provides reliable cluster of storage facility, filtered traffic data can be fetched easily and can provide end user with traffic analysis and useful predictions.
SIGNIFICANCE OF INFORMATION TECHNOLOGY IN HR DEPARTMENT
Dr. Manasi Bhate
Computer Department, MES IMCC, Savitribai Phule Pune University, Pune, India
Keywords: Information Technology (IT), Human Resource (HR).
ABSTRACT
Information technology deals with technology and other aspects of managing and processing information, especially in large organizations. Human Resource Development is the framework for helping employees to develop their personal and organizational skills, knowledge, and abilities. For analyzing the Implementation of IT in HR Department, we have conducted a survey with the help of questionnaire considering various objectives. Primary Data was collected from eight Information Technology organizations. After analysis of data, results were presented in the paper.
ENHANCED CERTIFICATELESS ID BASED AUTHENTICATION SYSTEM
V.Isakkirajan, Research Scholar, Manonmaniam Sundaranar university, Tirunelveli 627012 virpknc@gmail.com,
Dr. M.Ramakrishnan Professor & Head, Department of Computer Applications,
Madurai Kamaraj University, Madurai.- 625021
ramkrishod@mkuniversity.org
Keywords: Certificateless, Identity, Signcryption, Encryption, Decryption, Authentication.
ABSTRACT
Certificateless encryption does not require the use of certificate to guarantee the authencticity of public keys. It does not rely on the use of Trusted Centre (TC) who is in possession of a master key. In these respects, We make concrete the concept of Certificateless authentication encryption by introducting identity based encryption, Signature and key exchange schemes. We demonstrate how identity based encryption can be supported to secured communication between two entities with Certificateless encryption. The Schemes are all implemented using RSA algorithm. The lack of certificates and the desire to prove the schemes secure in the presence of an adversary who has access to the public key requires the careful development of new secure model. In this paper, we prove that the Certificateless Identity Scheme is secure in a fully adaptive adversial model.
REVIEW OF MACHINE LEARNING ALGORITHMS FOR IT OPERATIONS
Dr. Santosh Deshpande¹, Rahul Gaikwad²
¹Director, MES’S Institute of Management and Career Courses Pune, India
²Sr. DevOps Engineer, FireEye and Research scholar, Savitribai Phule Pune University, Pune, India
Keywords: Log Analysis, Failure Prediction, Text Mining, Machine Learning Algorithms.
ABSTRACT
In IT operations, system alerts and failures are very difficult to predict but it’s becoming crucial for business. If we are able to predict alerts or system failures it would really help to prevent unexpected and critical system downtimes also it would also assures reliability and high availability of business services for end users. Depends on system architecture , there are several types of logs generates like Application logs, database logs, network logs, server and business process logs etc. These logs are footprint of any activity performed by the user or any internal system activity. Logs includes critical and huge amount of information. In most of production environment, huge amount of logs produced per second. As human being , it’s not feasible to read these logs line by line . To resolve any system alerts or failures , there is strong need of a mechanism which can go analyse all these logs, events and alerts and provide more insights. By applying Machine learning algorithms we can automatically analyze and process these logs and alerts to get more insights which helps to understand and reduce the system downtimes. This paper focuses on analysis of different machine learning algorithms for IT system logs and alerts with their proposed outcomes in detail.
SURVEY ON MALWARE DETECTION IN ANDROID PHONE
Prajakta Jadhav¹, Shraddha Ghorpade², Prof. M. K. Kumbhar³
JSPM’s Bhivrabai Sawant Institute Of Technology and Research
Department of Information Technology1,2,3
Keywords: Android malware, Machine learning, Keywords Correlation Distance, SVM.
ABSTRACT
With the popularity of Android devices, more and more Android malware are manufactured every year. How to filter out malicious app is a serious problem for app markets. Analysing applications in order to identify malicious ones is a current major concern in information security; In view of the traditional feature extraction method based on binary program, this paper presents a method for feature extraction of JAVA source code. The method uses the Keywords Correlation Distance to compute the correlation between key codes such as API calls, Android permissions, the common parameters, and the common key words in Android malware source code. Then SVM is applied to make the system gain to accommodate the function of the new malicious software sample, so as to detect new malicious software and existing malwares.
OPTIMIZED SEMANTIC RETRIEVAL OF TEXT AND DESIGN
TRADEMARK BASED ON CONCEPTUAL SIMILARITY
1Prof.Pramod Dhamdhere, 2Roshni Kantale, 3Snehal Machale, 4Nikita Pardeshi, 5Vaishali Warkhade
Department of Information Technology and Engineering,
JSPM’s Bhivrabai Sawant Institute of Technology and Research Wagholi, Pune 412207.
Keywords: Conceptual similarity, trademark infringement, trademark retrieval, trademark similarity.
ABSTRACT
A trademarks is a sign that you can use to distinguish your business goods or services from those of other traders. Trademark can be defined expressly in the form of any symbol, logo, titles etc. so, they need to be secure. This project deciphers the hypothetic similarities among trademarks, which happens when more than two or more trademarks hail equal or relevant semantic implant. The state-of-the-art by offering a semantic algorithm to similitude trademarks in pre conditions of hypothetic parallelism. By using data similarity, it is derived that search and indexing technique developed similarity distance. The offered reflow algorithm is confirmed using two resources: a trademark database of conflicting cases and a databases company names. Extends the conceptual model by developing and evaluating a semantic algorithm for trademark retrieval based on conceptual similarity. The conceptual comparison of text documents that share similar domain, use similar concepts, or express similar ideas has been studied extensively. The underlying technology embedded in existing trademark search systems is primarily based on text-based retrieval. Use the different domains to measures the accuracy of the algorithm which gathered different data.
SMART AGRICULTURE SYSTEM USING IoT
Shraddha Udage, Gauri Shende, Shweta Dharmadhikari, Jayashree Pajai, and Prof.S.P.Godse
Sinhgad Academy of Engineering Pune, India, Pune 412207.
Keywords: Sensors, agricultural productivity, crop production, prediction, Internet of Things.
ABSTRACT
IoT advances can be use in smart forming to enhance quality of agriculture. Agriculture, backbone of Indian economy. A system consist of various features like Authentication at the user level, sensors, server based remote monitoring system, humidity and temperature sensing, soil and moisture sensing, remote control motor means automated water uses monitoring system in effective way etc. The propose system developed on information send from sensors and estimated the quality of water needed for enhance the productivity. The major advantage of the system implementing with cloud computing, that will optimize the selection crop of particular field also uses of water fertilizers while maximizing the yield of crops. The concept of IoT is used to continuously monitor and track water usage via wireless sensor nodes. Server collect the data through WiFi to process, track usages and wastage of water.
FARMER ADVISORY SYSTEM FOR CROP
CULTIVATION USING SPEECH RECOGNITION TECHNOLOGY
Poonam Mhetre1, Kajal Tiwari2, Utkarsha Bhandwalkar3, Prof. Premalatha G.4
1,2,3,4Department of Computer Engineering, Imperial College Of Engineering and Research Wagholi, Pune, India
Keywords: Data mining, speech recognition, Android SDK, client module.
ABSTRACT
Farmer Advisory System is useful for farmer in solving their queries related to farm cultivation. We develop system through which user can ask their questions by submitting it in the form of voice command as well as in textual format. There are different Farm Management Developer available but all these systems use 2-tier architecture . In this, only client read information which is already stored in database. There is no interaction between user and system. We develop a Farm Management System for this we use 3 tier architecture. It is an interactive application. In this we provide a speech recognition facility. User can ask their queries in the form of audio commands. The system converts this voice into text format. Then this, retrieve information from the database which is related to query. After retrieving data, it display to users. Query which are related to farm such as crop, land, soil, fertilizers, expenditures, etc. and this information store in database.
IMPLEMENTATION OF HIGH ACCURACY-BASED IMAGE
TRANSFORMATION MODULE IN CLOUD COMPUTING
A Nanda Gopal Reddy¹, Dr. B. V. Ramana Murthy²
¹Research Scholar, Department of CSE, Rayalaseema University, Kurnool, AP, India
²STANLEY College of Engineering & Technology, Hyderabad, India
(E-mail: nandagopalreddy@gmail.com)
Keywords: Digital Image Processing; Healthcare Domain; Cloud Computing; component.
ABSTRACT
Several different algorithms are significantly developing day by day in the field of digital image processing this technique advancement is growing rapidly with suitable trends for the next generation. This application are mainly used in the field of healthcare and hospital management system and these application are implemented in the real-time scenario by cloud computing. This paper presents the structure of comparative strategically algorithm in each of the phases through a sequential procedure of different phases of acquisition transformation interpolation, filtering, edge detection and image recognition in the cloud computing .We capture the data flow process in each of this phase and analysis the data flow and transformation process in each of the phase It also shows in each and every stage how the data has been read and transformed from one phase to another phase are discussed. This paper summarizes the emerging techniques of the next generation is the digital image process the photography techniques in this field is regularly changing according to the trends and technology advancement.
IOT BASED NAVIGATION ROBOT
Prof. P.G. Patil, Prasad Javharkar, Manoj Dashpute, Akshay Damale, Sanju Kheriya
Sandip Institute of Technology & Research Centre, Savitribai Phule Pune University, Nashik, India
Keywords: Log Analysis, Failure Prediction, Text Mining, Machine Learning Algorithms.
ABSTRACT
In IT operations, system alerts and failures are very difficult to predict but it’s becoming crucial for business. If we are able to predict alerts or system failures it would really help to prevent unexpected and critical system downtimes also it would also assures reliability and high availability of business services for end users. Depends on system architecture, there are several types of logs generates like Application logs, database logs, network logs, server and business process logs etc. These logs are footprint of any activity performed by the user or any internal system activity. Logs includes critical and huge amount of information. In most of production environment, huge amount of logs produced per second. As human being , it’s not feasible to read these logs line by line. To resolve any system alerts or failures , there is strong need of a mechanism which can go analyse all these logs, events and alerts and provide more insights. By applying Machine learning algorithms we can automatically analyze and process these logs and alerts to get more insights which helps to understand and reduce the system downtimes. This paper focuses on analysis of different machine learning algorithms for IT system logs and alerts with their proposed outcomes in detail.
ATTRIBUTE-BASED STORAGE SUPPORTING SECURE
DEDUPLICATION OF ENCRYPTED DATA IN CLOUD
USING DROP TECHNIQUE
Miss. Swati Dhokate and Prof. N. M. Shivale
Department of Computer Engineering, JSPM’s BSIOTR, Wagholi, Pune-412207
Keywords: ABE, De-dupllication, storage.
ABSTRACT
In public cloud storage system protecting the data and controlling the data access is a challenging issue. Cipher text Policy Attribute-Based Encryption (CP-ABE) has been adopted as a promising technique to provide flexible, fine-grained and secure data access control for cloud storage with honest-but-curious cloud servers. However numerous works have been proposed using CP-ABE scheme, in which the single attribute authority must execute the time-consuming user legitimacy verification and secret key distribution and hence it results in a single-point performance bottleneck when a CP-ABE scheme is adopted in a large-scale cloud storage system. Clients may be stuck in the waiting line for a long stretch to get their mystery keys, which results in low-efficiency of the framework. Even though the multi authority access control plans have been proposed, these plans still cannot conquer the disadvantages of single-point bottleneck and low efficiency; because of the way that each of the authority still autonomously deals with a disjoint characteristic set. In order to overcome this disadvantage, there has been proposed a novel heterogeneous framework to remove the problem of single point performance bottleneck and provide a more efficient access control scheme with an auditing mechanism. This framework employs multiple attribute authorities to share the load of user legitimacy verification. Meanwhile, in this scheme, a CA (Central Authority) is introduced to generate secret keys for legitimacy verified users and each of the authorities in this scheme manages the whole attribute set individually. This system makes performance improvement in key generation and also guarantees security requirement. Still there are some security loopholes in this system such as there is no protocol to verify owner and If the owner is compromised then he/she may put wrong data or information in the data server and users may get wrong data. There is no way to know who has used the data.
CREDIT RISK ANALYSIS USING MACHINE LEARNING ALGORITHMS
Harsh Kantawala¹, Brijesh Patel²
¹Department of Computer Engineering, G. H. Patel College of Engineering and Technology, Gujarat Technical University, India
²Assistant Professor, Department of Computer Engineering, G. H. Patel College of Engineering and Technology, India
Keywords: Support Vector Machines, Neural Network, Credit Risk, Loan Classification, Decision Tree, Classification.
ABSTRACT
The allowing of loans by a Bank is one of the important decision problems that require care. It can be performed by a different kind of processing. This Theory describes a credit evaluation system that uses different machine learning algorithms. We will train and implement five different algorithms to take decision about to approve or reject the loan application. Credit scoring is the famous technique in credit risk analysis which has an active research area in financial risk management. Here, in this analysis process we will take 1000 applicant details with categorical and numerical attributes. The data in numerical form will be used in neural network. Machine learning algorithms with different training-to-validation data ratios have been investigated, and a comparison between their implementation results has been checked. Experiment will give the result about the machine learning algorithms that are used in which learning scheme, can the proposed credit risk evaluation system deliver optimum performance; where it may be used efficiently, and quickly in automatic processing of credit applications.
STATISTIC ANALYSIS IN DATA MINING USING HADOOP ANALYSIS
Shradha Modak, Prof. G. M. Kadam
SKN Sinhgad Institude of Technology & Science, India
Keywords: Mining Enterprises; Technical and Economic data; BP Neural Network; Prediction Models.
ABSTRACT
The analysis method of the technical data and economic data is researched using technologies of big data analysis and data mining . The sales price data of mineral products is an important economic indicator of the geological data and mining enterprises is an important technical data. Characteristics of the technical and economic data of mining enterprises are nonlinearity and multi-dimensionality. In this we successfully analyzed the influencing factors and the fluctuation pattern. The results show that the prediction accuracy is high and the prediction model is strong. The prediction model of the mineral products price is established using artificial neural ne0074work. During the process of mineral development, the accuracy of the orebody shape is reduces, due to the limitation of technical conditions and equipment conditions, lots of geological data have been lost. The prediction model of the geological missing data is established using the techniques of geostatistics and artificial neural network. By using the model, the regularity of geological data of single borehole, the regularity of geological data of group boreholes and the regularity of geological data of all boreholes is discussed and analyzed.It has been proved that most of the geological missing data can be predicted and interpolated and the result of interpolation and prediction are reliable.
LIGHTWEIGHT SHAREABLE AND TRACEABLE SMART HEALTHCARE SYSTEM
Asst. Prof. S. S. Deshmukh1, Madhuri Pawar2, Rutuja Lodhi3, Pooja Mawale4, Punam Patil5
1Assistant Professor, 2,3,4,5UG Student, Dept. of Computer Engineering, ICOER, Pune
Keywords: Access control, searchable encryption, tractability, user revocation, mobile health system.
ABSTRACT
Portable wellbeing (m-Health) has developed as another patient driven model which permits continuous accumulation of patient information by means of wearable sensors, collection and encryption of this information at cell phones, and afterward transferring the encoded information to the cloud for storage and access by human services staff and scientists. In any case, proficient and adaptable sharing of encoded in-formation has been an extremely difficult issue. In this paper, we propose a Lightweight Sharable and Traceable (LiST) secure versatile wellbeing framework in which tolerant information are scrambled end-to-end from a patient’s cell phone to information clients. Rundown empowers productive catchphrase hunt and fine-grained get to control of encoded information, underpins following of double crossers who offer their look and access benefits for money related pick up, and permits on-request client denial. Rundown is lightweight as in it offloads the majority of the substantial cryptographic calculations to the cloud while just lightweight operations are performed toward the end client gadgets. We formally characterize the security of LiST and demonstrate that it is secure without irregular prophet. We likewise direct broad examinations to get to the framework’s execution.
ANNUAL CROP YIELD PREDICTION AND RECOMMEND PLANTING OF DIFFERENT CROPS BY USING DATA MINING TECHNIQUES
Naushina Farheen Imam Shaikh¹, Prof. R. V. Argiddi²
¹PG Student, Walchand Institute of Technology, Solapur
²Assistant Professor, Walchand Institute of Technology, Solapur
Keywords:
ABSTRACT
The agricultural assembly innovation is an innovation to make high esteem generation productivity, quality increment of farming items in the entire procedure of agricultural creation. Likewise, executing accuracy horticulture, this is an option in contrast to the future farming, through the union innovation that permits forecast of free market activity, ongoing administration and quality support amid the whole life cycle of agricultural items. Agribusiness is one of the significant wellsprings of survival and a standout amongst the most significant factors in the monetary development of the nation. Scientists in the field of the agriculture have considered and executed distinctive efficient systems which would foresee and expand the harvest yield sand make farming very beneficial.
SMART INFERENCE MODEL FOR IOT USING SVM
Jadhav Dipali J.1, Kadam Gayatri S.2, Ghodke Poonam S.3, Prof. Pokharkar S. T.4
¹UG Student, Dept. of Computer Engineering, Rahuri Factory
²Assistant Professor, Dept. of Computer Engineering, Rahuri Factory
Keywords: Log Analysis, Failure Prediction, Text Mining, Machine Learning Algorithms.
ABSTRACT
The requirement for brilliant IOT applications is developing as IOT is turning into a vital piece of our everyday life. Web of-Things (IoT) have effectively associated a great many gadgets to the Internet and expanding each day. As the nation is moving towards digitalization, the proposed keen IOT based framework encourages the requirement for induction of colossal information that is produced by the IOT based gadgets. So we thought of structuring a task which will diminish the abundance measure of information created by the framework and sent to the cloud. A keen induction channel will be planned which will remove the undesirable information that are to be sent to the cloud servers. AI calculations will be actualized to diminish the measure of information exchanged brilliantly by different parameters. Our framework will have a blend of Wi-Fi, Smart telephone, Machine learning and cloud to accomplish our objective.
ANALYSIS OF DOCKER SYSTEM RESOURCE UTILIZATION
Dr.Nagaraja G.S¹, Jayanth N²
¹Professor, Department of Computer Science and Engineering,
R.V College of Engineering, Affiliated to VTU, Bangalore 560059, India
²Backend Intern Engineer, Department of Computer Science and Engineering,
R.V College of Engineering, Affiliated to VTU, Bangalore 560059, India
Keywords: Docker Container, Virtual mahine, Docker images, Network nodes, Container networking.
ABSTRACT
Docker is used to ship the application. It has many components like Docker hub, Registry, image and Cloud service. Manually Network Nodes are deployed in both virtual machine and container. This is done to check two conditions, one is to check the reduction in time complexity and another is to check space complexity problem. Image is build before container is created. Time taken by the virtual machine will be always high and dependency rate is also be on higher side. Space taken or memory consumption by the container to run the build is on lower side compared to virtual machine. This hypothesis is shown by running some real nodes inside the docker and checking the dependency rate. Time taken for the ‘n’ number of nodes can be seen in lower side in container and the alternate hypothesis of container taking less time can be achieved.
MOBILE CONTROLLED WIRELESS BASED APPROACH FOR HUMAN ROBOT INTERACTION
Meenakshi Panchal¹, Prof. Shailesh Jadhav²
¹PG Student, Dept of E&TC, DPCOE, Pune
²Assistant Professor, Dept of E&TC, DPCOE, Pune
Keywords: IOT, WIFI-Module, NodeMCU, Android app, Robotic arm.
ABSTRACT
Robots are increasingly being integrated into working tasks to replace humans to perform the repetitive task. Generally, robotics can be divided into two areas, industrial and service robotics. When we talk about robots, people tend to think that robots are only use in the industry or just for the scientist to test about new technologies. The robot body was prepared mechanically and electrical components were chosen to be suitable to be used as a robotic arm. This project presents the development of a Mobile controlled wireless based robotic arm, Mobile Android App consists of provision for receiving 3 different inputs that are voice, hand gesture, phone tilt.
EFFICIENT SEARCH SCHEME OVER ENCRYPTED DATA ON MOBILE CLOUD
Seema Kadam¹, Prof. S. M. Rokade²
¹PG Student, Dept. of Computer Engineering, Pravara Rural Engineering College, Loni
²Assistant Professor, Dept. of Computer Engineering, Pravara Rural Engineering College, Loni
Keywords: Mobile cloud storage, searchable data encryption, energy efficiency, traffic efficiency.
ABSTRACT
Cloud storage provides a convenient, massive, and scalable storage at low cost, but data privacy is a major concern that prevents users from storing files on the cloud trustingly. One way of enhancing privacy from data owner point of view is to encrypt the files before outsourcing them onto the cloud and decrypt the files after downloading them. However, data encryption is a heavy overhead for the mobile devices, and data retrieval process incurs a complicated communication between the data user and cloud. Normally with limited bandwidth capacity and limited battery life, these issues introduce heavy overhead to computing and communication as well as a higher power consumption for mobile device users, which makes the encrypted search over mobile cloud very challenging. In proposed system, we propose TEES (Traffic and Energy saving Encrypted Search), a bandwidth and energy efficient encrypted search architecture over mobile cloud. The proposed architecture offloads the computation from mobile devices to the cloud, and we further optimize the communication between the mobile clients and the cloud. It is demonstrated that the data privacy does not degrade when the performance enhancement methods are applied. Our experiments show that TEES reduces the computation time and save the energy consumption on file retrieval, meanwhile the network traffics during the file retrievals are also significantly reduced.
REVIEW FORECASTING EVENTS OVER SOCIAL MEDIA USING CHRONO-SPATIAL MODEL
Dr. Ankit Kabra¹, Prof. M. P. Wankhade²
¹PG Student, Dept. of Computer Science Engineering, SCOE, Pune
²Assistant Professor, Dept. of Computer Science Engineering, SCOE, Pune
Keywords: User Topic Participation (UTP), Spatio-Temporal Model, Topic Re-hotting Prediction.
ABSTRACT
To distinguish hotly debated issues is incredibly outstanding which can benefit various endeavors including topic suggestions, the guidance of public opinions, and so forth. Individuals might need to know when to re-hot a subject in any case, at times i.e., make the topic famous yet again. By showing a Spatio-Temporal User Topic Participation (UTP) display which models users’ practices of posting messages proposed system address this issue in this task. In online social networks the UTP exhibit thinks about users’ interests and friend circles and surprising events. Similarly, since subjects are changing reliably after some time, it considers the determined spatio-temporal displaying of topics. To smooth the differences in theme re-hotting prediction a weighting scheme is proposed. Finally, preliminary outcomes coordinated on obvious enlightening accumulations display the viability of the proposed models and topic re hotting prediction methods.
REMOVAL OF RECORD DUPLICATION FROM MULTIPLE SOURCES
Shubhangi Pansare¹, Prof. Nawathe A. N.²
¹PG Student, Dept. of Computer Engineering, Amrutvahini College Of Engineering, Sangamner
²Assistant Professor, Dept. of Computer Engineering, Amrutvahini College Of Engineering, Sangamner
Keywords: Record normalization, records fusion, web information integration.
ABSTRACT
Record normalization from multiple sources is very difficult problem. The promise of Big Data depends upon addressing numerous massive facts integration challenges, such as report linkage at scale, actual-time information fusion, and integrating Deep Web. In this paper, we formalize the record normalization problems, present in-intensity evaluation of normalization granularity degrees (e.g., record, field, and value-component). Our goal is to generate the unique, standard record for each cluster of true matching records, we call the resulted record as normalized record also we proposed a Strategy that includes both single strategy and multi-strategy approaches. Single strategy approaches are frequency, length, centroid, and feature-based to select the normalized data or for multi strategy approach, we used result merging models to combine the results from a number of single strategies. We will conduct experiments on the conference related data and the experimental results verify the effectiveness of our design by means of comparing with the present strategies.
LEVERAGING DEDUPLICATION IN ENCRYPTED CLOUD DATA WITH ATTRIBUTE SUPPORT
Sayali Milind Aher¹, Prof. S. D. Jondhale²
¹PG Student, Dept. of Computer Engineering, Pravara Rural Engineering College, Loni
²Assistant Professor, Dept. of Computer Engineering, Pravara Rural Engineering College, Loni
Keywords: ABE, De-dupllication, storage.
ABSTRACT
In public cloud storage system protecting the data and controlling the data access is a challenging issue. Cipher text Policy Attribute-Based Encryption (CP-ABE) has been adopted as a promising technique to provide flexible, fine-grained and secure data access control for cloud storage with honest-but-curious cloud servers. However numerous works have been proposed using CP-ABE scheme, in which the single attribute authority must execute the time-consuming user legitimacy verification and secret key distribution and hence it results in a single-point performance bottleneck when a CP-ABE scheme is adopted in a large-scale cloud storage system. Clients may be stuck in the waiting line for a long stretch to get their mystery keys, which results in low-efficiency of the framework. Even though the multi authority access control plans have been proposed, these plans still cannot conquer the disadvantages of single-point bottleneck and low efficiency; because of the way that each of the authority still autonomously deals with a disjoint characteristic set. In order to overcome this disadvantage, there has been proposed a novel heterogeneous framework to remove the problem of single point performance bottleneck and provide a more efficient access control scheme with an auditing mechanism. This framework employs multiple attribute authorities to share the load of user legitimacy verification. Meanwhile, in this scheme, a CA (Central Authority) is introduced to generate secret keys for legitimacy verified users and each of the authorities in this scheme manages the whole attribute set individually. This system makes performance improvement in key generation and also guarantees security requirement. Still there are some security loopholes in this system such as there is no protocol to verify owner and If the owner is compromised then he/she may put wrong data or information in the data server and users may get wrong data. There is no way to know who has used the data.CA who generates secret keys, is assumed to be fully trusted, If CA gets compromised he can collude with any user or AA to provide secret keys to illegitimate users. In order to overcome these disadvantages, we are going to propose a new framework where there will be a protocol to verify owner and his/her data to be uploaded and a log will be maintained to know who will be accessing the data. This framework will also propose to choose one among the AAs to act as CA instead of separate CA and will have an observer to trace if CA is working properly or not. If observer finds any discrepancy it will be creating a report. This will make the system more secure and efficient.
SESPHR: A METHODOLOGY FOR SECURE SHARING OF PERSONAL HEALTH RECORDS IN THE CLOUD
Nikita Gholap¹, Prof. S. K. Korde²
¹PG Student, Dept. of Computer Engineering, Pravara Rural Engineering College, Loni
²Assistant Professor, Dept. of Computer Engineering, Pravara Rural Engineering College, Loni
Keywords: Access Control, Cloud Computing, Personal Health Records, Privacy.
ABSTRACT
The broad acknowledgment of cloud based services in the healthcare sector has brought about practical and helpful trade of Personal Health Records (PHRs) among a few taking part elements of the e-Health systems. Nevertheless, putting away the secret health data to cloud servers is susceptible to revelation or theft and requires the improvement of approaches that guarantee the protection of the PHRs. Along these lines; we propose an approach called SeSPHR for secure sharing of the PHRs in the cloud. The SeSPHR scheme ensures patient-centric control on the PHRs and preserves the classification of the PHRs. The patients store the encrypted PHRs on the un-trusted in cloud servers and specifically grant access to various kinds of clients on various parts of the PHRs. A semi-trusted in proxy called Setup and Re-encryption Server (SRS) is acquainted with set up people in general/private key combines and to deliver the re encryption keys. Besides, the strategy is secure against insider threats and furthermore authorizes a forward and in reverse access control. Moreover, we formally examine and check the working of SeSPHR strategy through the High Level Petri Nets (HLPN). Execution assessment in regards to time utilization demonstrates that the SESPHR strategy can possibly be utilized for safely sharing the PHRs in the cloud.
MULTI-LANE POTHOLE DETECTION FOR VEHICLE SENSOR DATA
Trupti Mhaske¹, Prof. S. M. Rokade²
UG Student, Department of Computer Engineering, PREC, Loni
Assistant Professor, Department of Computer Engineering, PREC, Loni
Keywords: GPS Receiver, Ultrasonic Sensor, Wireless Sensor Network, Micro-controller.
ABSTRACT
Pothole Detection framework is an exceptional idea and it is extremely valuable to whom which confront the issue of pothole in their route. The innovation is absolutely new and thought is created a profile for pothole in your vehicle venture. It is an application which is Accessing to auspicious and precise road condition data, particularly about hazardous potholes is of incredible significance to people in general and the government. System proposes a successful road surface observing framework for computerized pothole recognition. It is a special idea where it a low cost solution for the road safety purpose. This will maintain a strategic distance from accidents and can use to recognize issue regions early. The experts can be alarmed to take preventive activities; preventive activities can set aside extra cash. Ineffectively kept up road is an unavoidable truth in most creating nations including our India. A very much kept up road organize is an absolute necessity for the prosperity and the advancement of any nation. So we will make a viable road surface observing framework. Mechanized pothole recognition is our concentration in the framework.
COLLABORATIVELY SOCIAL IMAGE UNDERSTANDING USING DEEP MATRIX FACTORIZATION
Miss. Ashwini B. Shete¹, Prof. Anuradha N. Nawathe²
¹UG Student, Department of Computer Engineering, AVCOE, Sangamner
²Assistant Professor, Department of Computer Engineering, AVCOE, Sangamner
Keywords: Image Understanding, Tag Refinement, Tag Assignment, Image Retrieval.
ABSTRACT
The quantity of pictures related with pitifully administered client gave labels has expanded drastically lately. Client gave labels are insufficient, abstract what’s more, uproarious. In proposed framework, we center on the issue of social picture understanding, i.e., label refinement, label task, and picture recovery. Unique in relation to past work, system propose a novel feebly administered profound lattice factorization calculation, which reveals the dormant picture portrayals and label portrayals installed in the inert subspace by cooperatively investigating the feebly directed labeling data, the visual structure, and the semantic structure. The semantic and visual structures are mutually fused to take in a semantic subspace without over fitting the uproarious, deficient, or abstract labels. Additionally, to expel the loud or repetitive visual highlights, an inadequate model is forced on the change grid of the first layer in the profound design. Broad examinations on true social picture databases are led on the assignments of picture understanding: picture label refinement, task, and recovery. Empowering results are accomplished, which shows the adequacy of the proposed strategy.
SITUATIONAL ANALYTIC METHOD FOR USER BEHAVIOR PATTERN IN MULTIMEDIA SOCIAL NETWORKS
Kirti Pulate¹, Prof. S. D. Jondhale²
¹PG Student, Department of Computer Engineering, PREC, Pune
²Assistant Professor, Department of Computer Engineering, PREC, Pune2
Keywords: Multimedia Social Networks, Situation Analytics, Intention Prediction, Behavior Pattern, Mental Disorder Detection, Big Data.
ABSTRACT
In today’s world, it is undeniable that social media plays an important role in impacting our culture, our economy and our overall view of the world. Social media is a new forum that brings people to exchange idea, connect with, relate to, and mobilize for a cause, seek advice, and offer guidance. Most research on social network mining focuses on discovering the knowledge behind the data for improving people’s life. While Multimedia Social Networks (MSNs) seemingly expand their user’s capability in increasing social contacts, they may actually decrease the face-to-face interpersonal interactions in the real world. Therefore, the interaction behaviors between users and MSNs are becoming more comprehensive and complicated. This project primarily extended and enriched the situation analytics framework for the specific social domain and further proposed a novel algorithm for users intention serialization analysis based on classic Generalized Sequential Pattern (GSP). We leveraged the huge volume of user behaviors records to explore the frequent sequence mode that is necessary to predict user intention. Our experiment selected two general kinds of intentions: playing and sharing of multimedia, which are the most common in MSNs, based on the intention serialization algorithm under different minimum support threshold (Min Support). By using the users microscopic behaviors analysis on intentions, we found that the optimal behavior patterns of each user under the Min Support, and a users behavior patterns are different due to his/her identity variations in a large volume of sessions data. We also propose machine learning based, Social Network Mental Disorder Detection (SNMDD), which exploits features extracted from social network data to accurately identify potential cases of SNMDs so we can find out the stressed users on social media platforms.
ACTIVE USER PREDICTION, RANKING AND PROVIDING ADS IN SOCIAL NETWORKING SERVICES BASED ON USERS PROFILE
Mr. Nilesh Gholap¹ and Dr. Mininath Bendre²
¹Student, Pravara Rural Engineering College, India
²Associate Professor, Pravara Rural Engineering College, India
Keywords: Behaviour Pattern, Mental Disorder Detection, Multimedia Social Networks, Situation Analytics, Intention Prediction.
ABSTRACT
Social media presence has been predominant today, for example, twitter.com and facebook.com, where a huge number of users continue interacting with one another consistently. One interesting and essential issue in the user to user communication administrations is to rank users dependent on their activeness in an appropriate manner. In this paper, we propose a one of a kind point of view to accomplish this objective, which is measuring user activeness by examining the dynamic associations among users on social systems. In view of this thought, we create quantitative estimations for user activeness and propose our first algorithm for positioning the users. Likewise we further consider the common impact between users while figuring the important estimations and propose the second positioning algorithm, which registers user activeness in an iterative manner.
FACE DETECTION USING VIOLA-JONES ALGORITHM &
CONVOLUTIONAL NEURAL NETWORKS
Manav Bansal and Dr Rupak Sharma
Research Scholar of Monad University, Department of Ccomputer Science, Asst. Professor, Department of Computer Applications at SRM Institute of Science & Technology, U.P
Keywords: Face Detection, Face recognition, system, Viola-Jones.
ABSTRACT
In the event that you have ever utilized web based life, a computerized camera, or a mobile phone, odds are you have experienced face detection more than once. Well known online networking applications extending from Facebook to Snapchat use face detection for a few of their prevalent highlights, for example, labeling companions and applying channels. Face detection is characterized as PC innovation that is utilized to recognize human faces in computerized pictures. There are different PC algorithms that are utilized in the field of face detection, however this paper will concentrate on two of the most well known strategies: Convolutional Neural Networks and the Viola-Jones calculation. The inspiration for this paper is a general interest about face detection in regular daily existence just as an oddity about how face detection algorithms work, in my classes so far I have discussed the structure of various kinds of PC frameworks however I have never dug into algorithms for these frameworks. This task offered the chance to find out about face detection as well as dive into the universe of PC algorithms.
PROBLEMS AND CHALLENGES IN SMART VEHICLES: A SURVEY
Jissy Ann George¹, Priyanka Surendran¹
¹AMA International University, Kingdom of Bahrain
Keywords: Smart vehicles, Cybersecurity, Data collection, Parking data, Block chain, Encryption.
ABSTRACT
Smart vehicles have become an essential part of every growing country. As the population increases, the number of vehicles keep increasing as well. There are a number of challenges that have come up due to these. The number of IoT devices that will be connected will run to tens of billions by the year 2020. This will in turn cause a number of challenges such as security, maintainability, scalability, etc. Hence it is essential to identify the possible breaches that can occur when using smart vehicles.