Proposed Research Work Research Articles

An Examination on Systems in Place for Managing E-Waste with Procedures and Operations Involved in Collection, Treatment and Disposal of E-Waste

E-waste is a catch-all phrase that refers to a wide range of products that contain electric and electronic components. Old electrical and electronic equipment is referred to as "e-waste" or "e-waste". Computers, televisions, mobile phones, printers, and white electronic appliances like refrigerators are a few examples. It is well recognised that e-waste contains a variety of toxic or hazardous components that, if released during processing, recycling, or disposal, constitute a major risk to both human health and the environment. This proposed research work examines on systems in place for managing and collecting E-Waste with the procedures and operations involved in collection , treatment and disposal of E-Waste.

An intelligent disease prediction system for psychological diseases by implementing hybrid hopfield recurrent neural network approach

Machine learning based approaches for automatic disease prediction is a novel research area in healthcare informatics. Electronic Health Records in medical settings improves early-stage illness diagnosis. However, when standard rule-based approaches, like doctor's prescription or laboratory test reports are employed for disease diagnosis, the advantages of EHRs are not accomplished adequately. As a result, there is a requirement of technology based solution which helps in prediction of psychological diseases in a more efficient way. The proposed research work offers a hybrid Hopfield recurrent neural network (H2RN2) approach to predict psychological diseases by using amorphous clinical EHRs taken from Kaggle database. The proposed model automatically learns inherent semantic characteristics from available clinical data items. It uses fivefold cross validation technique within a recurrent neural network which detracts over fitting of the model. In addition to effective learning during training of the model, the hybrid approach also helps in accurate prediction of the disease with improved accuracy. The proposed model is assessed using three measuring parameters, accuracy, recall and F1-score and yields an accuracy of 97.53% in experimental evaluation, which is superior to several existing approaches for psychological disease prediction. The results demonstrate that the proposed model outperforms several other techniques in predicting the risk of psychiatric disorders. In future, the similar approach may be employed to predict gender-based psychological diseases or to anticipate the risk of various physiological diseases.

PREDICTION OF CAR RESELLING PRICES USING MACHINE LEARNING

It is generally known that, taking wise and challenging decisions is really a crucial task in every business. To propose a novel solution for this challenge, this research work majorly focuses on one of the retail businesses i.e., used car sales business. The proposed research work shows that, the predictive analytical models will be a great add- on to business mainly for assisting the decision making process. The major objective of our paper is to build a prediction model i.e., a fair price mechanism to predict the cars selling price based on their features like the car model, the number of years that a car is old, the type of fuel it uses, the type of seller, the type of transmission and the number of kilometers that the car has driven so far. This paper will help to get an approximation about selling price of a used car based on its features and reduces the seller and consumer risk in business. The proposed model utilizes the machine learning algorithms and regression techniques of statistics like linear, decision tree and random forest regressions to achieve this task. KEYWORDS Used car sales business, decision making process, machine learning algorithms, regression techniques, linear, decision tree and random forest regressions.

Automatic Recognition of Medicinal Plants: Based on Multispectral and Texture Features using Hidden Deep Learning Model

Identification of medicinal plants automatically in the environments is necessary to know about their existence around us. Recently, there are many techniques followed to recognize plants automatically such as through leaves and flowers with their shape and texture. Leaf-based plant species identification systems are widely used nowadays. This proposed research work uses a deep learning approach using Convolutional Neural Networks (CNN) to recognize medicinal plants through leaves with high accuracy. For this research, leaf images are collected from nature and used as the experimental dataset. The authors have collected leaf items from 5 different medicinal plants. After the collection of images and have to pre-process them which plays an important role in the classification steps. Deep learning model and algorithm are used for classification purposes among them, VGG16 worked pretty well and got an accuracy level of 95.48%. In real life, this paper can well affect the medical sector and learn more about medicinal plants.

IoT Supervised PV-HVDC Combined Wide Area Power Network Security Scheme Using Wavelet-Neuro Analysis

Power system networks are one of the most widely used methods in the real world for transferring large amounts of electrical energy from one location to another. At present, High Voltage Direct Current Transmission is preferred for long distances over hundreds of miles due to minimal power loss and transmission cost of transmission.Due to an increase in power demand, integration of renewable sources to minimise the voltage fluctuations and compensate for power loss is necessary. This is a mandatory requirement to produce sophisticated protection methods for mainly smart systems under various balanced and unbalanced fault conditions. The system protection scheme must respond as quickly as possible to protect the connected devices in a smart environment. The network must be monitored and protected under various weather conditions as well as electrical parametric problems. The proposed research work is carried on the basis of physical monitoring with the aid of the Internet-of-Things and electrical parameters calibrated with the help of wavelet analysis. A wavelet is a mathematical tool to investigate the behaviour of transient signals at different frequencies, which provides important information related to the detailed analysis of faults in power networks. The major goals of this research are to analyse faults using detailed coefficients of current signals through the bior-1.5 mother wavelet for fault identification and artificial neural network analysis for fault localization. This proposed approach furnishes an IoT supervised Photovoltaic - High Voltage Direct Current (HVDC) combined wide area power network security scheme using wavelet detailed coefficients under various types of faults with Fault-Inception-Angles.

Physiological Synchrony Predict Task Performance and Negative Emotional State during a Three-Member Collaborative Task.

Evaluation of team performance in naturalistic contexts has gained popularity during the last two decades. Among other human factors, physiological synchrony has been adopted to investigate team performance and emotional state when engaged in collaborative team tasks. A variety of methods have been reported to quantify physiological synchrony with a varying degree of correlation with the collaborative team task performance and emotional state, reflected in the inconclusive nature of findings. Little is known about the effect of the choice of synchrony calculation methods and the level of analysis on these findings. In this research work, we investigate the relationship between outcomes of different methods to quantify physiological synchrony, emotional state, and team performance of three-member teams performing a collaborative team task. The proposed research work employs dyadic-level linear (cross-correlation) and team-level non-linear (multidimensional recurrence quantification analysis) synchrony calculation measures to quantify task performance and the emotional state of the team. Our investigation indicates that the physiological synchrony estimated using multidimensional recurrence quantification analysis revealed a significant negative relationship between the subjectively reported frustration levels and overall task performance. However, no relationship was found between cross-correlation-based physiological synchrony and task performance. The proposed research highlights that the method of choice for physiological synchrony calculation has direct impact on the derived relationship of team task performance and emotional states.

Cow Milk Quality Grading using Machine Learning Methods

Milk is considered as complete food as it contains rich set of proteins and vitamins. Therefore determining quality of cow milk plays an important role in today’s research. In this paper four methods are implemented to check quality of cow milk using dataset consists of 1059 milk samples taken from various cows. Three grades of milk grade A, B, C are considered based on different features of cow milk. Various machine learning methods K Nearest neighbors, Logistic regression, Support Vector machine and ANN are implemented. Accuracy of these methods is then compared. It has been observed that the results of KNN (n=3) is more accurate amongst all four methods implemented in the proposed research work.

A sophisticated semantic analysis framework using an intelligent tweet data clustering and classification methodologies

Developing an efficient twitter sentiment analysis framework is one of the most challenging and demanding chores in current days. In the conventional works, various machine/deep learning based detection mechanisms are developed for tweet analysis. It encompasses the methodologies of data clustering, optimization, and classification, but it has the major problems of difficult to understand, complexity in mathematical computations, high time consumption, and error rate. Therefore, the proposed research work intends to implement an intelligent and advanced data mining techniques to design and develop a sentiment analysis framework. At first, the preprocessing is performed to normalize the dataset for generating the noise free balanced dataset. After that, the Hierarchical Tweet Expression Clustering (HTEC) algorithm is used to cluster the attributes according to the distance for simplifying the process of classification. Moreover, the multi-faceted features such as vector space Bag-of-Words (BOW), and Term Frequency – Inverse Document Frequency (IDM) are extracted for classifier training and testing operations. Finally, the Spatial Dense Bi-directional Long Short Memory (SDBi-LSTM) classification methodology is used for predicting the sentiment as positive, negative, or neutral. For evaluation, the Stanford Twitter Sentiment Test Set (STS-Test/Sentiment140), GOP debate and IMDB datasets are used to validate the results of the proposed system by using the parameters of accuracy, f1-score, training time, testing time, true positive rate, false positive rate, and etc. Also, the obtained results are compared with the recent prediction models for proving the superiority of the proposed framework.

Cluster-based segmentation for tobacco plant detection and classification

Tobacco is one of the major economical crops in the agriculture sector. It is essential to detect tobacco plants using unmanned aerial vehicle (UAV) images for improved crop yield and plays an important role in the early treatment of tobacco plants. The proposed research work is carried out in three phases: In the first phase, we collect images from UAV’s and apply the French Commision Internationale de l'eclairage (CIE) L*a*b colour space model as pre-processing operations and segmentation. And then two prominent motion descriptors namely histogram of flow (HOF) and motion boundary histogram (MBH) are combined with the optimal histogram of oriented gradients (HOG) descriptor for exploring optimal motion trajectory and spatial measurements. And finally, the spatial variations with respect to the scale and illumination changes are incorporated using the optimal HOG descriptor. Here both dense motion patterns and HOG are refined using hierarchical feature selection using principal component analysis (PCA). The proposed model is trained and evaluated on different tobacco UAV image datasets and done a comparative analysis of different machine learning (ML) algorithms. The proposed model achieves good performance with 95% accuracy and 92% of sensitivity.

Analysis on the aluminium dross waste as a base material for making brick for construction purposes

ABSTRACT The proposed research work is to develop the brick by incorporating steel industry industrial waste to achieve sustainable environment. Aluminium dross (AD) is the one of the major waste produced during aluminium ingots production. Particle size and chemical compositions of aluminium dross are analysed using Scanning Electron Microscopy (SEM). To effectively utilise the aluminium dross, five mix composition of bricks were produced to determine the influence of AD on mechanical properties of bricks such as the compressive strength, water absorption, efflorescence and bulk density. Mix B gives the maximum compressive strength of 11.54 N/mm2, water absorption of 5.6% and bulk density of 1658.4 kg/m3 for the mix of 5% dross. Analysis of variance (ANOVA) is employed to analyse the influence of dross and fly ash concentrations on the quadratic model.

Automated Brain Hemorrhage Classification and Volume Analysis

Brain hemorrhage is a serious and life-threatening condition. It can cause permanent and lifelong disability even when it is not fatal. The word hemorrhage denotes leakage of blood within the brain and this leakage of blood from capillaries causes stroke and adequate supply of oxygen to the brain is hindered. Modern imaging methods such as computed tomography (CT) and magnetic resonance imaging (MRI) are employed to get an idea regarding the extent of the damage. An early diagnosis and treatment can save lives and limit the adverse effects of a brain hemorrhage. In this case, a deep neural network (DNN) is an effective choice for the early identification and classification of brain hemorrhage for the timely recovery and treatment of an affected person. In this paper, the proposed research work is divided into two novel approaches, where, one for the classification and the other for volume calculation of brain hemorrhage. Two different datasets are used for two different techniques classification and volume. A novel algorithm is proposed to calculate the volume of hemorrhage using CT scan images. In the first approach, the ‘RSNA’ dataset is used to classify the brain hemorrhage types using transfer learning and achieved an accuracy of 93.77%. Furthermore, in the second approach, a novel algorithm has been proposed to calculate the volume of brain hemorrhage and achieved tremendous results as 1035.91 mm3 and 9.25 cm3, using the PhysioNet CT scan tomography dataset.

NUMERICAL SOLUTION OF WAVELET NEURAL NETWORK LEARNING WEIGHTS USING ACCELERATED PARTICLE SWARM OPTIMIZATION ALGORITHM

Artificial Neural Network (ANN) is a supervised learning nonlinear complex model. This characteristic enables ANN to be used in nonlinear system modeling and classification applications. This research work proposed a technique called Wavelet Neural Network (WNN), and in the hidden layers of WNN, Morlet and Mexican are used as an activation function. During the processing, the WNN gets stuck in the local minimum causing slow convergence. For evaluating such kinds of problems, numerous algorithms have been tried and used. Consequently, this proposed research work used a novel meta-heuristic search technique called Accelerated Particle Swarm Optimization (APSO) algorithm combined with the WNN. Due to the effective convergence and fast searching toward an optimal solution, the APSO algorithm is used. In the proposed APSOWNN algorithm, APSO searches for the best sub-search solution. In conclusion, this model is assessed on the basis of total of three different datasets like the 4-bit OR, 7-bit Parity and IRIS benchmark classification problems, and its efficiency is equated with criterion methods such like Wavelet Back Propagation Neural Network (WBPNN), Artificial Bee Colony Wavelet Neural Network (ABCWNN) and WNN. Finally, from the results of simulation, it has been concluded that the proposed algorithm’s performance is much better, as compared to the state-of-the-art algorithms in terms of mean square error.(MSE) and accuracy.

MICROWAVE IRRADIATION SYNTHESIS OF NITROGENDOPED Ag/MWCNT/ZnO NANOCOMPOSITE FOR PHOTOCATALYTIC APPLICATION

In this proposed research work, Zinc oxide is fabricated with carbon nanotubes, nitrogen, and silver. The conclusion of the distinctive hybrid of dopants on the ZnO/MWCNT nanocomposite was studied. Acidic and basic nature dyes are used to check the efficiency of the composite photocatalyst. The hydrothermal & microwave route was adapted for combining mesoporous nanocomposite samples. The XRD analysis confirmed the information to the nanoscale crystal with the hexagonal dimension of ZnO. The observation of UV-vis spectra validates its photocatalytic competency in the visible beam as an outcome of the reduction in the bandgap energy. The proficiency of the photocatalyst was additionally entrenched over the axing in electron-hole recombination activity referring to dwindling PL intensity. The BET inspection supports the mesoporous structures of the samples. The SEM micrograph revealed the existence of agglomeration in nanoparticles with MWCNT while the perusal of TEM images established the existence of a 16–24 nm diameter of the MWCNT. Furthermore, the EDX analysis confirmed the presence of the Zn, N, O, Ag, and C.

Precise Multi-Class Classification of Brain Tumor via Optimization Based Relevance Vector Machine

The objective of this research is to examine the use of feature selection and classification methods for distinguishing different types of brain tumors. The brain tumor is characterized by an anomalous proliferation of brain cells that can either be benign or malignant. Most tumors are misdiagnosed due to the variability and complexity of lesions, which reduces the survival rate in patients. Diagnosis of brain tumors via computer vision algorithms is a challenging task. Segmentation and classification of brain tumors are currently one of the most essential surgical and pharmaceutical procedures. Traditional brain tumor identification techniques require manual segmentation or handcrafted feature extraction that is error-prone and time-consuming. Hence the proposed research work is mainly focused on medical image processing, which takes Magnetic Resonance Imaging (MRI) images as input and performs preprocessing, segmentation, feature extraction, feature selection, similarity measurement, and classification steps for identifying brain tumors. Initially, the median filter is practically applied to the input image to reduce the noise. The graph-cut segmentation technique is used to segment the tumor region. The texture feature is extracted from the output of the segmented image. The extracted feature is selected by using the Ant Colony Optimization (ACO) algorithm to improve the performance of the classifier. This probabilistic approach is used to solve computing issues. The Euclidean distance is used to calculate the degree of similarity for each extracted feature. The selected feature value is given to the Relevance Vector Machine (RVM) which is a multi-class classification technique. Finally, the tumor is classified as abnormal or normal. The experimental result reveals that the proposed RVM technique gives a better accuracy range of 98.87% when compared to the traditional Support Vector Machine (SVM) technique.

Signal Conducting System with Effective Optimization Using Deep Learning for Schizophrenia Classification

Signal processing based research was adopted with Electroencephalogram (EEG) for predicting the abnormality and cerebral activities. The proposed research work is intended to provide an automatic diagnostic system to determine the EEG signal in order to classify the brain function which shows whether a person is affected with schizophrenia or not. Early detection and intervention are vital for better prognosis. However, the diagnosis of schizophrenia still depends on clinical observation to date. Without reliable biomarkers, schizophrenia is difficult to detect in its early phase and hence we have proposed this idea. In this work, the EEG signal series are divided into non-linear feature mining, classification and validation, and t-test integrated feature selection process. For this work, 19-channel EEG signals are utilized from schizophrenia class and normal pattern. Here, the datasets initially execute the splitting process based on raw 19-channel EEG into 6250 sample point’s sequences. With this process, 1142 features of normal and schizophrenia class patterns can be obtained. In other hand, 157 features from each EEG patterns are utilized based on Non-linear feature extraction process where 14 principal features can be identified in terms of considering the essential features. At last, the Deep Learning (DL) technique incorporated with an effective optimization technique is adopted for classification process called a Deep Convolutional Neural Network (DCNN) with mayfly optimization algorithm. The proposed technique is implemented into the platform of MATLAB in order to obtain better results and is analyzed based on the performance analysis framework such as accuracy, Signal to Noise Ratio (SNR), Mean Square Error, Normalized Mean Square Error (NMSE) and Loss. Through comparison, the proposed technique is proved to a better technique than other existing techniques.

DC–DC Converter with Pi Controller for BLDC Motor Fuzzy Drive System

The Brushless DC Motor drive systems are used widely with renewable energy resources. The power converter controlling technique increases the performance by novel techniques and algorithms. Conventional approaches are mostly focused on buck converter, Fuzzy logic control with various switching activity. In this proposed research work, the QPSO (Quantum Particle Swarm Optimization algorithm) is used on the switching state of converter from the generation unit of solar module. Through the duty cycle pulse from optimization function, the MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) of the Boost converter gets switched when BLDC (Brushless Direct Current Motor) motor drive system requires power. Voltage Source three phase inverter and Boost converter is controlled by proportional-integral (PI) controller. Based on the BLDC drive, the load utilized from the solar generating module. Experimental results analyzed every module of the proposed grid system, which are solar generation utilizes the irradiance and temperature depends on this the Photovoltaics (PV) power is generated and the QPSO with Duty cycle switching state is determined. The Boost converter module is boost stage based on generation and load is obtained. Single Ended Primary Inductor Converter (SEPIC) and Zeta converter model is compared with the proposed logic; the proposed boost converter achieves the results. Three phase inverter control, PI, and BLDC motor drive results. Thus the proposed grid model is constructed to obtain the better performance results than most recent literatures. Overall design model is done by using MATLAB/Simulink 2020a.

A Predictive Model for Abnormal Conditions in Smart Farming using IoT Sensors

By combining sensors and devices, the Internet of Things (IoT) technology revolutionizes smart farming by enabling real-time monitoring and data-driven decision-making. However, challenges persist, particularly in managing crop diseases and unusual weather patterns. Innovative solutions, such as neural network-based modules, provide early warning systems, empowering farmers to address these concerns proactively, and ensuring sustainable farming practices. The proposed research work incorporates a customized neural network module to detect unusual crop diseases and abnormal weather patterns, acting as an early warning system. The proposed technology delivers critical insights for timely interventions, allowing farmers to be proactive. Farmers can increase crop output and encourage sustainable agricultural methods by quickly detecting these uncommon challenges. This technique not only improves agricultural output, but it also improves the resilience and durability of farming ecosystems, assuring agriculture's long-term viability. The simulation results demonstrated that the proposed model outperforms other state-of-the-art smart farming models.

Grey Wolf Optimizer Based Deep Learning for Pancreatic Nodule Detection

At an early point, the diagnosis of pancreatic cancer is mediocre, since the radiologist is skill deficient. Serious threats have been posed due to the above reasons, hence became mandatory for the need of skilled technicians. However, it also became a time-consuming process. Hence the need for automated diagnosis became mandatory. In order to identify the tumor accurately, this research proposes a novel Convolution Neural Network (CNN) based superior image classification technique. The proposed deep learning classification strategy has a precision of 97.7%, allowing for more effective usage of the automatically executed feature extraction technique to diagnose cancer cells. Comparative analysis with CNN-Grey Wolf Optimization (GWO) is carried based on varied testing and training outcomes. The suggested study is carried out at a rate of 90%–10%, 80%–20%, and 70%–30%, indicating the robustness of the proposed research work. Outcomes show that the suggested method is effective. GWO-CNN is reliable and accurate relative to other detection methods available in the literatures.

Performance evaluation of LSB steganography based on multiple image formats

In the proposed research work steganography technique has been designed and developed for carrier images having different formats. The algorithm is applicable for the image formats: PNG (Portable Network Graphics), TIFF (Tagged Image File Format), JPEG (Joint Photograph Expert Group), and BMP (Bitmap). A single algorithm for almost all types of image formats provides many advantages. Depending upon the size of the secret message, network bandwidth, and allowable distortion we can choose the format of the carrier. For embedding secret messages, the LSB substitution technique is applied in this paper. We utilized the concept of imperceptibility by calculating PSNR (Peak Signal to Noise Ratio), MSE (Mean Square Error), SSIM (Structure Similarity Index Metrics), Histogram, and Entropy. The proposed work is analyzed for robustness with favorable results. Two steganalysis tools: Stegexpose and Stegdetect are used to detect the hidden secret message. For analyzing the complexity of work capacity, concealing and retrieval times are calculated. Furthermore, the comparative outcome for the proposed work is very promising for different carrier image formats.

Computer-Aided Diagnosis Model Using Machine Learning for Brain Tumor Detection and Classification

The Brain Tumor (BT) is created by an uncontrollable rise of anomalous cells in brain tissue, and it consists of 2 types of cancers they are malignant and benign tumors. The benevolent BT does not affect the neighbouring healthy and normal tissue; however, the malignant could affect the adjacent brain tissues, which results in death. Initial recognition of BT is highly significant to protecting the patient’s life. Generally, the BT can be identified through the magnetic resonance imaging (MRI) scanning technique. But the radiotherapists are not offering effective tumor segmentation in MRI images because of the position and unequal shape of the tumor in the brain. Recently, ML has prevailed against standard image processing techniques. Several studies denote the superiority of machine learning (ML) techniques over standard techniques. Therefore, this study develops novel brain tumor detection and classification model using met heuristic optimization with machine learning (BTDC-MOML) model. To accomplish the detection of brain tumor effectively, a Computer-Aided Design (CAD) model using Machine Learning (ML) technique is proposed in this research manuscript. Initially, the input image pre-processing is performed using Gaborfiltering (GF) based noise removal, contrast enhancement, and skull stripping. Next, mayfly optimization with the Kapur’s thresholding based segmentation process takes place. For feature extraction proposes, local diagonal extreme patterns (LDEP) are exploited. At last, the Extreme Gradient Boosting (XGBoost) model can be used for the BT classification process. The accuracy analysis is performed in terms of Learning accuracy, and the validation accuracy is performed to determine the efficiency of the proposed research work. The experimental validation of the proposed model demonstrates its promising performance over other existing methods.