VOLUME 9, ISSUE 7, JULY 2020

Abstract: This paper proposed an ensemble hierarchical model to combine two or more classifiers which has been trained independently, and then fused them in the next level. This is done in two steps, first we trained a Decision Tree and a Logistic Regression models, step two we feed the output of those models to a Neural Network. The Neural Network is also trained to combine the output of previous classifiers to achieve better overall accuracy. To test our hypothesis, we used PIMA Indian diabetes database as benchmark problem. Our proposed model has achieved classification accuracy above 83% which is better than other states of the art methods in the literature. Keywords: Data mining, Regression, Neural Network, Decision Tree, Pima Diabetes Data set, Ensemble Learning.

Abstract: The project is intended to classify the Brain MRI using the computer vision technique of Deep Learning. Brain MRI can be of two major types depending on the way they were extracted from the scanner as well as on the type of scanner and method used for taking the MRI of a subject. These two categories are T1weighted anatomical as well as T2weighted resting BOLD MRI images. These are categorized with the help of a Deep Neural Networks with 4 hidden layers. The dataset is taken from the openfmri. There are 120 MRI data and are released to the general public on as a part of the materials for “Temporal interpolation alters motion in fMRI scans: magnitude and consequence for artifacts detection” by Power et al. in PLOS ONE. Encased for each subject could be a T1-weighted anatomical picture (MP-RAGE) and one or extra T2*-weighted scans(resting bold outputs).The dataset we have is a 3D cuboid of the subject’s MRI image for the T1 weighted scans and 4D for the T2 weighted scans where the 1st, 2nd and 3rd being the x, y and z axis for the 3D image and the 4th dimension being the time. Every subject’s MRI is then split into 2D slices from all the axis to increase the data volume, then these images are pre-processed and fed into a 2D-CNN network. This is then trained for 3 epoch cycle in the cloud for a better processing speed and the resulted output of the weighted and biases are stored for the model to predict future inputs.

Keywords: MRI, BOLD, T1-weighted, T2-weighted, 2D-CNN.

Abstract: As food being the major survival for human beings, the quality of food plays a vital role, so agriculture becomes the base for all food ingredients. The fruits or vegetables that are obtained from the farm must have a better quality. The quality depends on the techniques used for the effective growth of the crops. There are various techniques which are used to monitor the growth of each plant. One such technique is the Internet of Things, which connects things and people together. This helps the farmers to produce good products and also consumers to have good healthy food. This paper proposes a method in which sensors are used to monitor the growth of onions by implementing smart farming to produce maximum yield and healthy onions.

Keywords: Internet of things, Sensors, Onion, Weather Monitoring, Soil Monitoring, Crop Monitoring, Fertilizers.

Abstract: Agricultural productivity is something on which economy highly depends. The detection of plant leaf disease is a very important factor to prevent serious outbreak. Detection of plant disease through some automatic technique is beneficial as it reduces a large work of monitoring in big farms of crops, and at very early stage itself it detects the symptoms of diseases. Most plant diseases are caused by fungi, bacteria, and viruses. Thus, we propose a system for image segmentation technique which is used for automatic detection and classification of plant leaf diseases.

Keywords: Image Segmentation, Automatic Detection and Agricultural productivity.

Abstract: The rising number of kidney failure in adolescents and young children is of great concern. Pediatric CKD is a dynamic and complex medical and psychosocial disease with unique factors that separate this population from adults. Due to the unique and complex physical, psychological, and family backgrounds, young children may develop damage of kidneys. The long-term mortality for children, adolescents, and young adults with CKD (Chronic Kidney Disease) remains substantially higher than their healthy counterparts. The complex challenges that adolescent and young adult CKD patients face has to be dealt with on a serious note. Adolescents have different CKD etiologies and progress are quite dissimilar to that faced by adults, but have similar multifarious comorbidities. CKD can delay and limit growth. In this paper, various Machine Learning algorithms are used to predict the occurrence of the disease. The benefit of implementing this technique is that the disease can be diagnosed at an early stage based on the various symptoms of the patient and thus can help them to get the diagnosis and treatment on time which will lead to better health and better Quality of Life. Here, the prediction skill of several machine-learning algorithms for early prediction of CKD has been analyzed by usage of predictive analytics, in which the association of data parameters and the target class attributes is done. Predictive analytics enables us to introduce the optimal subset of parameters to feed machine learning to build a set of predictive models.

Keywords: Adolescents, Chronic Kidney Disorder, Machine Learning Algorithms.

Abstract: In the future, smart machines will update or enhance human competencies in many areas. Artificial intelligence is the intelligence shown by machines or software. Artificial Intelligence is becoming a popular area in laptop science, as it has enhanced human life in many areas. Artificial intelligence in the last decades has dramatically improved the overall performance of the manufacturing and service systems. Study in the region of artificial intelligence has given rise to the rapidly developing technology regarded as an expert gadget. Application regions of Artificial Intelligence are having a significant impact on numerous fields of lifestyles as the professional gadget is broadly used in recent times to remedy the complicated troubles in diverse regions as science, engineering, business, medicine, weather forecasting. The regions using the era of Artificial Intelligence have visible growth in excellent and efficient. This paper offers an overview of this technology and the utility areas of this era. This paper will also discover the frequent use of Artificial Intelligence technologies in the PSS design to damp the electricity device oscillations as a result of interruptions, in Network Intrusion for protecting computer and conversation networks from intruders, within the precise location- medicine, to improve hospital inpatient care, for medical picture type, within the accounting databases to mitigate the issues of it.

Keywords: Intrusion Detection Systems, Artificial Intelligence, Neural Networks (computer), Power System Stabilizer.

Abstract: In the present study we present an innovative approach towards countering the problem of partial occlusion in face recognition scenario. The partial occlusion can be caused by various objects such as scarfs, sunglasses etc., and its effects are confounding in the performances of the recognition rates. The advantage that the adopted pre-processing algorithm poses before face recognition steps is to eliminate the distortions due to the variance in light illumination field at the given instance when the facial image is recorded or captured. The framework tends to mathematically model the curvature and other essential features of the face such as micro-expression and the curves of the facial regions. This, significantly enhances the probability of matching the parent image to that of the occlude image that is how multiple object recognition using hybrid approach. The presented algorithm is tested over Extended Yale B & CMU PIE standardized datasets. Over the years biometrics has gained unparalleled popularity in digital medium and has proven its usefulness for several applications concerned with the threats and crime or security purposes. Face Recognition is a widely emerging biometric for automating the surveillance, as it has aid in strengthening the security from several types of terrorist or criminal threats. Though, there are several face recognition techniques which are categorized based on its error rates in recognition but there are few that gives the marginal rate for sufficient and validated recognition rates for occlude images.

Keywords: Hybrid Neuro Fuzzy Network, Face Recognition, Neural Network, Normalized Rapid Descriptor (NBD), PCA, LDA, NMF, LNMF, ICA.

Abstract: A research depicts that the average attrition rate in India is as high as 25%, as the employers fail to meet the expectations of the employee. On the other hand, another report states that the average attrition rate of employees in the Telecom, BFSI, aviation and financial services is about 31% and is tremendously higher in the IT/ITES sector. Enterprise culture is the soul of an enterprise, which is the key to obtain sustainable competitive advantage. For enterprise survival and development, enterprise culture is not the direct factor, but the most lasting decisive factor. In this paper, given the important role of enterprise culture in the process of human resource management practice, combining cultural construction with recruiting, training, utilizing, and retaining talent to improving the level of human resource management to achieve benign interaction between culture construction (of the company) and human resource management. An effort is made in realizing a long-term sustainable competitive goal to obtain an invincible position in the present competitive market. Human Resource Management can be defined as planning, organizing, directing and compensating human resources resulting in the creation and development of human relation with a view to contribute proportionately to the organizational and individual goals. The examination of raw or crude data and drawing conclusions out of it is called data analytics. In this paper we will be analysing the employee turnover pattern and the factors contributing to it. Efforts will be made to create a model that can predict if a certain employee will leave the company or not. The goal is to create or improve different retention strategies on targeted employees. The first step in data analytics- data pre-processing is presented in the paper. Data pre-processing techniques convert crude data into useful format. Real world data are generally incomplete- noisy, inconsistent and contains many errors. Removing these factors improves the quality of analysis and prediction. The focus of data analytics lies in inference, the process of deriving conclusions. In this paper 2 out of top 3 strategies affecting employee turnover are being analysed and graphs plotted. The 3 top features include evaluation v/s exit, average monthly income v/s exit and satisfaction v/s exit. In this paper we have taken up the challenge of predicting the exit vs. evaluation trends of the company, first using Logistic Regression method, and later using Random forest or Random decision forest method. The two models have to be compared at different instances of execution thereby, creating a platform of distinction between the two most popular Machine Learning algorithms today.

Keywords: Logistic Regression, Random decision forest, Human Resource Management, examination of raw or crude data and drawing conclusions- data analytics, employee turnover pattern, data pre-processing, evaluation v/s exit, average monthly income v/s exit, satisfaction v/s exit, planning, organizing, directing and compensating, competitive advantage.

Abstract: In order to make decision making effective, large amounts of unmined healthcare data are collected from the health care industry used to discover hidden information. These insights and their correlations are in most cases not used to their optimum, and thus paving way to more advanced data mining techniques. Medical diagnosis using machine learning is a complex task that requires humongous amounts of data that traditional decision support systems cannot provide answers to. For instance, the likelihood of patients being diagnosed with heart disease can be predicted by using medical profiles such as sex, blood pressure, and sugar level which enables the establishment of significant knowledge such as patterns that exacerbates chronic heart failures. The quest to solve these problems led to the development of a user-friendly web-based application on the Microsoft .NET framework to serve as a Clinical Decision Support System (CDSS) for cardiologist. Keywords: Machine Learning, K-Means, Naïve Bayes, Heart Disease, Clinical Support System

Keywords: Biometrics, Arduino, fingerprint

Abstract: Recently the buildings are constructed on large scale on a vast area. The building construction projects should be finished within the planned timeline and budget. However lots of the construction projects are delayed due to the many factors such as bad weather and delayed supply but main factor that affects the delay is the human error. Workers can misplace the equipment or skip the required job. Detecting the work error lately may cause delay and redoing the job. These errors should be detected as soon as possible to reduce the delay. But the monitoring and management and of construction work become difficult because the construction work is done in the very extensive area concurrently by lots of workers. To monitor the work progress, the construction company sends the project manager to the construction site and the managers recodes the change on to the paper sheet or tablet pc and append the photos. Later the company can view the progress by checking the paper or photo. Some companies these days use computer programs for the progress monitoring and the project manager can recode the change to the program and the program computes the progress. Recording the change in to the paper or computer program can check the work schedule but it is hard to detect the human error such as misplacement of equipment or skipping. In this paper we have proposed a system for construction site management where everything will be managed by this system which will help to manage construction site details systematically and it will save time and efforts. Keywords: CMS; cloud; templates; reports; media.

Abstract: Diabetes (Diabetes Mellitus), is a group of metabolic disorders and millions of people are affected. Detection of diabetes is of a great significance and serious complications should be concerned. Many research studies have been done on the diagnosis of diabetes, most of the research studies are based on one particular data set which is the Pima Indian diabetes data set. This Pima Indian data set is a data set of studies of women in India's population that began in 1965., and its onset rate is relatively high in diabetes. Most research studies were carried out prior to focusing primarily on one or two specialized complex techniques for testing data, while an inclusive research on several general techniques are missing. In this system, we extensively explore the most popular techniques in Machine Learning (e.g. KNN algorithm) used to identify the diabetes and pre-processing of data methods. We will examine this technique by the accuracy of the cross validation on the UCI ML repository data set.

Keywords: Machine learning, Classification, KNN, Diabetes.

Abstract: Future electric lights will be comprised of visible LEDs (light emitting diode).Visible LED’s with high power output are expected to serve in the next generation of lamps. An indoor visible data transmission system utilizing visible led lights is proposed. In the system, these devices are used not only for illuminating rooms but also for an optical wireless communication system. This system is suitable for private networks such as consumer communication networks. However it remains necessary to investigate the properties of visible LED’s when they are used as optical transmitters. Based on numerical analyses and computer simulations it can be used for indoor optical transmission. Infrared light is already used for communication, such as wireless remote control, IrDA, Infrared wireless LAN, and infrared inter-building communication. However, visible light LEDs are beginning to be used in every home and office, which makes visible light LEDs ideal for ubiquitous data transmitter.

Abstract: Fraud is the activity which will cause distress to corporations. This Financial fraud has been a huge worry for many organizations across the industries, this insurance industry comprises of over and above thousands companies all across the board, which gathers greater than a trillion dollars premium every year and billions of dollars are being lost every year because of this fraud, thus detection of Insurance Fraud is a burdensome task for the insurance companies. The conventional outlook for detecting insurance fraud was completely dependent on evolving heuristics around the fraudulent indicators. The auto insurance fraud is being considered to be one of the leading categories of fraud, which will be carried out by faking accident claim. In this study, we are concentrating towards tracking down the auto insurance fraud by making use of Machine Learning Techniques (Naïve Bayes Classifier), by using this techniques time complexity will be declined and also depicts the results accurately.

Keywords: Machine Learning techniques, Auto Insurance, Fraud detection, Naïve Bayes Classifier

Abstract: In the 21st century, we have observed a boom in the number of vehicles owned by people, but this sudden surge of automobiles led to the challenge of creating suitable parking and their maintenance. Although the parking of these vehicles was managed initially by manual labour, at this point of time it is a daunting task especially when the parking of these vehicles spans over multiple floors and every nook and cranny needs to be allocated specifically for the vehicles types and needs. With the emergence of electric vehicle all around the word in an effort to save the planet parking lots are being modified to accommodate charging ports for such vehicles and this would be a great time to inculcate additional features and controls to the parking to greatly reduce the reliance of humans and have a record of every activity happening inside the parking area. Although some limited number of parking lots have basic levels of electronic governance, an upgrade to fully monitored system is a much-needed change. As we have a responsibility to reduce wastage, it would also be wise to add on more features on top of the existing system without compromising efficiency while reducing the cost of products significantly. Therefore, choosing a modular approach is the wisest decision as when required more features could be added and older features could be removed from it without the need to remake the whole system. The main objective of this study is shed some light on the possibilities of improving such a system by adding more features and exploring possibilities for specified needs. The above mentioned would be aimed for while keeping in conditions like energy efficiency, cost efficiency and feasibility. In this study, the application would consider multiple real life factors using sensors like PIR(Passive infrared), Ultra-Sonic sensor, cameras etc and the compiled result would be used to help run the setup while notifying drives when needed using actuators like buzzer, lights etc. As the functionality of the application is unlimited, it would greatly benefit if it is connected to the internet, but even without the connectivity, it could be a self-reliant system which is able to perform its tasks in real time.

Keywords: Parking automation, IOT, Raspberry Pi, Grid System, Arduino, Modular.

Abstract: This article deals with pre-processing and enhancement activities such as removal of film artifacts and labels, filtering the image. A gradient based image enhancement method for mammography and MRI images is based on the first derivative and the local statistics. The proposed mammogram image enhancement method, the film artifacts such as labels and X-ray marks are removed and the high frequency components are removed using various filtering. The performance of the proposed method is also evaluated by means of Signal-to Noise-Ratio (SNR) using the mammographic images from the Mammography Image Analysis Society database and MRI brain images obtained from KOVAI Medical Center Hospital (KMCH), India. Keywords: MIAS, MRI, Artifacts, Filter, PSNR, ASNR

Abstract: Spiral inductors on semiconductor/insulator substrate play a crucial role in Radio Frequency Integrated Circuits (RFICs). For high frequency system circuitry, these components are realized using bond-wires or planar spirals. The Quality Factors (Q) of bond wires is higher than on-chip spirals, their use is constrained by the limited range of realizable inductances. The research presented here provides insight into some of the most pressing issues currently being addressed by the research community, and provides guidelines for designing these evolving heterogeneous Three Dimension (3D) systems. 3D spiral inductor integration is an evolving technology that will enhance the semiconductor roadmap for several generations. This dissertation provides insight into the 3-D inductor IC design process, with the goal of strengthening the design capabilities for 3-D integrated circuits and systems. Keywords: Inductor, SiO₂, Silicon, Glass, PEC.  

Abstract: Oral health helps to maintain the health state of all the structures like lips, teeth, gum, tongue and palate, good oral hygiene emphasis on cleanliness and moisturizing of mouth structures. Objectives of the study is to assess the prevalence of oral problems among school children, determine the awareness of oral hygiene among school children, develop and distribute information booklet regarding prevalence of oral problem and awareness regarding oral hygiene, co-relate the prevalence of oral problems with awareness of oral hygiene among school children,  find out the association between prevalence of oral problems and awareness of oral hygiene with selected sociodemographic variable. Material and Methods: Non- experimental descriptive research approach was employed descriptive design. A set of self-administered knowledge questionnaires was used to collect data. Written permission has obtained from the research ethical committee and formal written permission has also been obtained from the Principal of selected senior secondary schools The reliability of the tool was determined by using split half method and the tool was found to be highly reliable. Result: The findings reveal that 0.7% children had inadequate knowledge regarding oral hygiene, 7.0% children had moderate knowledge regarding oral hygiene, 92.3% children had adequate knowledge regarding oral hygiene. Mean Percentage Scores 46.02 and SD 6.118.

Keywords: Assess, Awareness, Oral Hygiene, Prevalence, School Children