Browse Items (2150 total)
Sort by:
-
Drought PredictionA Comparative Analysis of Supervised Machine Learning Techniques
Drought is a natural phenomenon that puts many lives at risk. Over the last decades, the suicide rate of farmers in the agriculture sector has increased due to drought. Water shortage affects 40% of the world's population and is not to be taken lightly. Therefore, prediction of drought places a significant role in saving millions of lives on this planet. In this research work, six different supervised machine learning (SML) models namely support vector machine (SVM), K-nearest neighbor (KNN), decision tree (DT), convolutional neural networks (CNNs), long short-term memory (LSTM), and recurrent neural networks (RNNs) are compared and analyzed. Three dimensionality reduction techniques principal component analysis (PCA), linear discriminant analysis (LDA), and random forest (RF) are applied to enhance the performance of the SML models. During the experimental process, it is observed that RNN model yielded better accuracy of 88.97% with 11.26% performance enhancement using RF dimensionality reduction technique. The dataset has been modeled using RNN in such a way that each pattern is reliant on the preceding ones. Despite the greater dataset, the RNN model size did not expand, and the weights are observed to be shared between time steps. RNN also employed its internal memory to process the arbitrary series of inputs, which helped it outperform other SML models. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Predicting Price Direction of Bitcoin based on Hybrid Model of LSTM and Dense Neural Network Approach
Bitcoin is a rapidly growing but extremely risky cryptocurrency. It marks a watershed moment in the history of cash. These days, digital currency is preferred to actual money. Bitcoin has decentralized authority and placed it in the hands of its users. Many people are joining the largest and most well-known Bitcoin mining pools as the risk of working alone is too great. In order to enhance their chances of creating the next block in the Bitcoins blockchain and decrease the mining reward volatility, users can band together to form Bitcoin pools. This tendency toward consolidation may also be seen in the rise of large-scale mining farms equipped with powerful mining resources and speedy processing capability. Because of the risk of a 51% assault, this pattern shows that Bitcoin's pure, decentralized protocol is moving toward greater centralization in its distribution network. Not to be overlooked is the resulting centralization of the bitcoin network as a result of cloud wallets making it simple for new users to join. Because of the easily hackable nature of Bitcoin technologies, this could lead to a wide range of security vulnerabilities. The proposed approach uses normalization and filling missing values in preprocessing, PCA for feature Extraction and finally training the model using LSTM-DNN Models. The proposed approach outperforms other two models such as CNN and DNN. 2023 IEEE. -
Optimal Disassembly Sequence Generation Using Tool Information Matrix
Just as the assembly sequence plays an important role in the early part of the product, the disassembly sequence plays an important part in the final stage of the product. The disassembly sequence determines how efficiently the product can be recycled or it can be disassembled for maintenance purposes. In this study, the disassembly sequence is generated using the Tool Information Matrix (TIM) and the contact relations. In this study the feasible sequences are generated using the TIM and contact relations, afterward, the time required is considered as a fitness equation for generating the optimal disassembly sequence. The proposed methodology is applied to 10-part crankshaft assembly to test the performance in generating the optimal disassembly sequences. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Exploring Bio Signals for Smart Systems: An Investigation into the Acquisition and Processing Techniques
Bio signals play a vital role in terms of communication in the absence of normal communication. Bio signals were automatically evolved from the body whenever any actions took place. There are lots of different types of bio signal based research going on currently from several researchers. Signal acquisition, processing the signals and segmenting the signal were totally different from one technique to another. Placing electrodes and its standard measurements were varied. The signals gathered from each subject may be varied due to their involvement. Each and every trial of signals can generate different patterns. Each and every pattern generated from the activities also has a different meaning. In this study we planned to analyze the basic measurement techniques handled to record the bio signals like Electrooculogram. 2023 IEEE. -
Simulation of the Electrical Control Unit (ECU) in Automated Electric Vehicles for Reliability and Safety Using On-Board Sensors and Internet of Things
The adaptation of the energy storage system (ESS) with high power and energy density remains a difficulty for electric vehicles (EVs), despite the increasing demand they are experiencing around the world. A lightweight, compact ESS is necessary to deliver the responsive performance and driving range that modern vehicles need. When planning for widespread use of EVs, it's important to give careful attention to the factors of ESS selection, sizing, and administration. One of the most promising future mobility alternatives is the hybrid electric vehicle (HEV), which offers improved fuel economy and lower pollution levels. As a result, one of the most pressing needs is for automakers to develop new technologies for vehicle design that might help lessen emissions and boost economy. The environmental impact of emissions from light-duty cars is growing in tandem with the annual increase in the number of such vehicles on the road. The usage of other modes of transportation, such as ships and planes, is on the rise, but road transportation will always be the most common. Electronic Control Units, or ECUs, have been increasingly commonplace in cars during the past few decades. Vehicle network multicore CPU scheduling is notoriously difficult. This study's findings consist of a straightforward power-sharing control approach for the HESS based on battery and UC, with the goal of extending the battery's useful life in a city environment. 2023 IEEE. -
Lightweight Anti DDoS Security Tool: Edge Level Filtering in SDN using P4
Software Defined Network (SDN) which has a promising future in satellite communication was first introduced as the solution to solve problems existing in the traditional network architecture. So far in SDN, mitigation strategies employed hardware installation or software solution which is heavily dependent on SDN controllers. The disadvantage of these approaches is the a) cost for implementation, b) intensive resource usage, and 3) costly optimization strategy necessary to enhance SDN performance. This research aims to fill the gap of the previously seen defense mechanism by enabling edge-level filtering without involving the control plane. By implementing filtering functions in edge switches, it can provide an efficient and effective defense layer in SDN network systems so that SDN switch can become the first line of defense against packet injection attacks. The proposed solution, Lightweight Anti-DDoS Software (LADS) focuses on lightweight workloads and provisioning of effective filtering mechanism to allow SDN switches to drop and block malicious packets sent by attackers. It utilizes Programming Protocol-independent Packet Processors (P4) programming language to create custom functionalities in SDN switches. P4 allows SDN switches to conduct host authentication and malicious packet filtering as well as blacklisting to isolate attackers. Simulation result proves that LADS efficiently manages malicious activities and maintains network performance during attacks at the data plane independent of SDN controller. 2023 IEEE. -
A Framework for Dress Code Monitoring System using Transfer Learning from Pre-Trained YOLOv4 Model
Maintaining a proper dress code in organizations or any environment is very important. It not only imbibes a sense of discipline but also reflects the personality and qualities of people as individuals. To follow this practice, some organizations like educational institutions and a few corporations have made it mandatory for the personnel to maintain proper attire as per their regulations. Manual checks are performed to adhere to the organizations' regulations which becomes tedious and erroneous most of the times. Having an automated system not only saves time but also there is very little scope of mistakes and errors. Taking this into context, the main aim and idea behind the project is to propose a model for detecting the dress code in such workplaces and educational institutions where the attire needs to be regularly monitored. The model detects Business Formals (Blazer, Shirt & Pants) worn by the personnel, for which CNN has been considered, along with YOLOv4, for performing the detection, due to its nature of giving the highest accuracy in comparison to the other object-detection models. Providing the Mean Average Precision of around 81%, it becomes evident that the model performs quite well in performing the detections. 2023 IEEE. -
Machine Learning Methods for Online Education Case
Online education has become a popular choice for learners of all ages and backgrounds due to its accessibility and flexibility. However, providing personalized learning experiences for a diverse range of students in online education can be challenging. Machine learning methods can be used to provide personalized learning experiences and improve student engagement in online education. In this case study, We're going to do some research on machine learning. methods in an online education platform. The platform provides courses in various subjects and is designed to be accessible to students from all over the world. The platform collects data on student behavior, such as the courses they enroll in, the time they spend on each course, and their performance on assignments and quizzes. We will explore several machine learning methods that can be applied to this data, including clustering, classification, and recommendation systems. Clustering algorithms can be used to group students based on their learning behavior and preferences, allowing instructors to provide personalized feedback and course recommendations. Classification algorithms can be used to predict student success in a particular course, allowing instructors to intervene and provide additional support if needed. Recommendation systems can be used to suggest courses to students based on their interests and past behavior. We will also discuss the potential benefits and challenges of using machine learning methods in online education. Benefits include increased student engagement, improved learning outcomes, and more efficient use of resources. Challenges include ensuring data privacy and security, preventing algorithmic bias, and maintaining transparency and fairness in the decision-making process. Overall, machine learning methods have the potential to transform online education by providing personalized learning experiences and improving student outcomes. By leveraging the vast amounts of data generated by online education platforms, we can create more effective and efficient learning experiences that meet the needs of students from diverse backgrounds and learning styles. 2023 IEEE. -
PMFRO: Personalized Mens Fashion Recommendation Using Dynamic Ontological Models
There is a thriving need for an expert intelligent system for recommending fashion especially focusing on mens fashion. As it is an area which is neglected both in terms of fashion and modelling intelligent systems. So, in this paper the PMFRO framework for mens recommendation has been put forth which indicates the semantic similarity schemes with auxiliary knowledge and machine intelligence in a very systematic manner. The framework intelligently creates mapping of the preprocessed preferences and the user records and clicks with that of the items in the profile. So, this model aggregates community user profiles and also maps the mens fashion ontology using strategic semantic similarity schemes. Semantic similarity is evaluated using Lesk similarity and NPMI measures at several stages and instances with differential set thresholds and the dataset is classified using the feature control, machine learning bagging classifier which is an ensemble model in order to recommend the mens fashion. The PMFRO framework is an intelligent amalgamation and integration of auxiliary knowledge, strategic knowledge, user profile preferences as well as machine learning paradigms and semantic similarity models for recommending mens fashion and overall precision of 94.68% and FDR of 0.06 was achieved using the PMFRO model. 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
Secure IBS Scheme for Vehicular Ad Hoc Networks
Vehicular Ad hoc Networks (VANET) havedrastically grown in recent years since they provide a better and more secure driving experience. Due to its characteristics, it is vulnerable to many security attacks. Even though many authentication schemes are proposed, their overheads are high. Hence, this study proposes a new Identity-Based Signature (IBS) for authentication with privacy-preservation. It supports secure communications with additional security features. It requires less overhead since it uses XOR operations and one-way hash functions for the signing and verification process. When the proposed schemes performance is compared to the recent schemes, it is observed that the proposed approach is more efficient in computation and communication. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Maximum Decision Support Regression-Based Advance Secure Data Encrypt Transmission for Healthcare Data Sharing in the Cloud Computing
The recent growth of cloud computing has led to most companies storing their data in the cloud and sharing it efficiently with authorized users. Health care is one of the initiatives to adopt cloud computing for services. Both patients and healthcare providers need to have access to patient health information. Healthcare data must be shared and maintained more securely. While transmitting health data from sender to receiver through intermediate nodes, intruders can create falsified data at intermediate nodes. Therefore, security is a primary concern when sharing sensitive medical data. It is thus challenging to share sensitive data in the cloud because of limitations in resource availability and concerns about data privacy. Healthcare records struggle to meet the needs of security, privacy, and other regulatory constraints. To address these difficulties, this novel proposes a machine learning-based Maximum Decision Support Regression (MDSR)-based Advanced Secure Data Encrypt Transmission (ASDET) approach for efficient data communication in cloud storage. Initially, the proposed method analyzed the node's trust, energy, delay, and mobility using Node Efficiency Hit Rate (NEHR) method. Then identify the efficient route using an Efficient Spider Optimization Scheme (ESOS) for healthcare data sharing. After that, MDSR analyzes the malicious node for efficient data transmission in the cloud. The proposed Advanced Secure Data Encrypt Transmission (ASDET) algorithm is used to encrypt the data. ASDET achieved 92% in security performance. The proposed simulation result produces better performance compared with PPDT and FAHP methods. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Weighted Mask Recurrent-Convolutional Neural Network based Plant Disease Detection using Leaf Images
Large losses in output, money, and quality/quantity of agricultural goods are incurred due to plant diseases. Seventy percent of India's GDP is tied to the agricultural sector, thus protecting plants from diseases is crucial. For this reason, it is important to keep an eye on plants from the moment they sprout. The usual approach for this omission is naked eye inspection, which is more time-consuming, costly, and requires significant skill. Thus, automating the method for detecting diseases is necessary to speed up this process. It is imperative that image processing methods be used in the creation of the illness detection system. Disease detection involves a number of processes, including Weighted Mask R-CNN, GLCM feature extraction, Multi-thresholding image pre-processing, and K means image segmentation classification. The weighted Mask R-CNN outperforms the standard RNN, the Mask R-CNN, and the CNN in terms of accuracy and recall in analytical trials by a significant margin. 2023 IEEE. -
Innovative Method for Detecting Liver Cancer using Auto Encoder and Single Feed Forward Neural Network
Liver cancer ranks sixth among all cancers in frequency of incidence. A CT scan is the gold standard for diagnosis. These days, CT scan images of the liver and its tumor can be segmented using deep learning and Neural Network techniques. In this proposed approach to identifying cancer cells, it's focus on four important areas: To enhance a photo by taking out imperfections and unwanted details. An ostu method is used for this purpose. Specifically, this proposed approach to use the watershed segmentation technique for image segmentation, followed by feature extraction, in an effort to isolate the offending cancer cell. After finishing the model training with AE-ELM. To do this, Extreme Learning Machine incorporates an auto encoder. To achieve effective and supervised recognition, the network's strengths of Extreme Learning Machine (ELM) are thoroughly leveraged, including its few training parameters, quick learning speed, and robust generalization ability. The auto encoder-extreme learning machine (AE-ELM) network has been shown to have a respectable recognition impact when the sigmoid activation function is used and the number of hidden layer neurons is set to 1200. According to the results of this investigation, a method based on AE-ELM can be utilized to detect the liver tumor. As compared to the CNN and ELM models, this technique achieves superior accuracy (around 99.23%). 2023 IEEE. -
TAMIL- NLP: Roles and Impact of Machine Learning and Deep Learning with Natural Language Processing for Tamil
Reading information in your mother tongue gives the feeling of enjoying juice of fruit. Researchers are working on regional languages to provide convenient and perfect automated tools to convert the content of knowledge from other languages. There exist many challenges based on the grammar of language. One of the classic regional languages, Tamil which is rich in Morphology, contains more processing challenges. The Natural Language Processing (NLP) technique along with Machine Learning (ML) and Deep Learning (DL) algorithms have been used to overcome those challenges. The accuracy of work is depending on the corpus provided to train the model. Among the reviewed papers using Support Vector Machine (SVM) of ML produced higher accuracy then other ML techniques. As DL techniques for NLP are booming one the researchers are working with different DL algorithms. Most of the NLP with Review Discussion in this paper will direct the researchers doing NLP in Tamil language to move further and to choose the right Machine Learning and Deep Learning algorithm to come out with accurate outcomes. 2023 IEEE. -
An Efficient Deep Learning Framework FPR Detecting and Classifying Depression Using Electroencephalogram Signals
Depression is a common and real clinical disease that has a negative impact on how you feel, how you think, and how you behave. It is a significant burdensome problem. Fortunately, it can also be treated. Feelings of self-pity and a lack of interest in activities you once enjoyed are symptoms of depression. It can cause a variety of serious problems that are real, and it can make it harder for you to work both at home and at work. The main causes include family history, illness, medications, and personality, all of which are linked to electroencephalogram (EEG) signals, which are thought of as the most reliable tools for diagnosing depression because they reflect the state of the human cerebrum's functioning. Deep learning (DL), which has been extensively used in this field, is one of the new emerging technologies that is revolutionizing it. In order to classify depression using EEG signals, this paper presents an efficient deep learning model that allows for the following steps: (a) acquisition of data from the psychiatry department at the Government Medical College in Kozhikode, Kerala, India, totaling 4200 files; (b) preprocessing of these raw EEG signals to avoid line noise without committing filtering; (c) feature extraction using Stacked Denoising Autoevolution; and (d) reference of the signal to estimate true and all. According to experimental findings, The proposed model outperforms other cutting-edge models in a number of ways (accuracy: 0.96, sensitivity: 0.97, specificity: 0.97, detection rate: 0.94). 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
Application of Artificial Intelligence in the Supply Chain Finance
Artificial intelligence (AI) has numerous applications in supply chain finance, including the ability to streamline processes, improve decision-making, and reduce costs. This abstract will discuss some of the key ways in which AI is being used in supply chain finance. One major Using AI in the Supply Chain finance is in risk management. By analyzing data from a variety of sources, including historical transaction data and external market data, AI can identify potential risks and suggest strategies for managing them. For example, AI can be used to predict which suppliers are at the greatest risk of financial distress, allowing companies to take proactive measures to minimize the impact of any disruptions. Another key Using AI in the Supply Chain finance is in fraud detection. By analyzing large volumes of data in real-time, AI can spot deviations from the norm that may point to fraud. This can help companies to prevent fraud and minimize losses. AI can also be used to optimize working capital management. By analyzing data on inventory levels, order volumes, and payment terms, AI can help companies to optimize their cash flow and improve their working capital position. For example, AI can help companies to identify opportunities to negotiate more favorable payment terms with suppliers or to optimize their inventory levels to minimize the amount of cash tied up in inventory. Finally, AI can be used to improve supply chain efficiency and reduce costs. By analyzing data on order volumes, shipping times, and other factors, A.I. may aid businesses in identify opportunities to their supply network needs improvement processes and reduce costs. For example, AI can aid businesses in determining opportunities to consolidate shipments or to optimize their routes to reduce transportation costs. Now a days AI has numerous applications in supply chain finance, including risk management, fraud detection, working capital management, and supply chain optimization. By leveraging the power of AI, companies can improve their financial performance, reduce costs, and enhance their overall competitiveness. 2023 IEEE. -
KMetaTagger: A Knowledge Centric Metadata Driven Hybrid Tag Recommendation Model Encompassing Machine Intelligence
The emergence of Web 3.0 has left very few tag recommendation structures compliant with its complex structure. There is a critical need for newer novel methods with improved accuracy and reduced complexity for tag recommendation, which complies with the Web 3.0 standard. In this paper, we propose KMetaTagger, a knowledge-centric metadata-driven hybrid tag recommendation framework. We consider the CISI dataset as the input, from which we identify the most informative terms by applying the Term Frequency - Inverse Document Frequency (TF-IDF) model. Topic modeling is done by Latent Semantic Indexing (LSI). A heterogeneous information network is formalized. Apart from this, the Metadata generation quantifies the exponential aggregation of real-world knowledge and is classified using Gated recurrent units(GRU). The Color Harmony algorithm filters out the initial feasible solutions into optimal solutions. This advanced solution set is finalized into the tag space. These tags are recommended along with the document keywords. When the suggested KMetaTagger's performance is compared to that of baseline techniques and models, it is found to be far superior. 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
A study of CNN models for re-identification of vehicles
Vehicle Re-identification has evolved in recent times. Initially, clicking a single picture of a vehicle or a car was done manually, inviting the workforce to complete a specified task. With the growth in technology, the method and techniques in Vehicle Re-Id also have advanced, transforming from manual to automation. Surveillance cameras were used to capture vehicle images and retrieve information about a specific vehicle. Re-trieving and identifying the images of the vehicle is done using computer vision, the most important branch of computer science and artificial intelligence. Earlier, Vehicle Re-Id implemented a single algorithm on a dataset, making the corresponding result insufficient to determine its effects. This paper proposes a brief survey of multi-modal techniques and methods for vehicle re-identification and fingerprinting. The different attributes of the vehicle are considered for ANPR (Automatic number plate recognition) for identifying the number plate, focusing on the vehicle's details or features as the initial phase of identification, and then the vehicle number plate. 2023 IEEE. -
Building Robust FinTech Applications and Reducing Strain on Strategic Data Centers using the LoTus Model
Agile is a well-known project management approach that has been used for many years. It places a strong emphasis on client satisfaction, adaptability, and teamwork. Agile was first developed as a software development approach, but it has now been modified for application in other sectors including marketing and finance. The Agile Manifesto, which was released in 2001 and explains the principles and ideals of Agile development, is the foundation of the Agile ideology. One or more of the guiding principles is to adapt to change instead of following a plan, prioritize functional software over thorough documentation, and collaborate with customers over negotiating contracts. Agile has gained popularity over time as businesses try to be adaptable and responsive to their customers' constantly changing business demands. The lack of predictability in Agile is one of its key drawbacks. Agile stresses client cooperation and adaptation, therefore the finished product could differ somewhat from what was originally planned. For businesses that depend on meticulous planning and a rigid schedule, this lack of predictability can be problematic. It faced a serious problem during the process of building a finance application called JazzFinance. This has led to build another robust and systematic software development method called as LoTus model. The proposed LoTus is an acronym for two abbreviations. Those are lean optimization TypeFace for Unified Systems (LoTus) and Locate dependencies, optimize for reusability, Test-Driven environment, Unify Design and Scalability. This article goes through the development of LoTus and how it has helped us build a stable finance application within a small amount of stipulated time. 2023 IEEE. -
ANN Based MPPT Using Boost Converter for Solar Water Pumping Using DC Motor
The solar DC pump system is simple to set up and run completely on its own without the need for human intervention. Solar DC pumps require fewer solar panels to operate than AC pumps. Solar PV Arrays, a solar DC regulator, and a DC pump make up the Solar DC Pump system. The nonlinear I-V characteristics of solar cells, PV modules have average efficiency compare to other forms of energy, and output power is affected by solar isolation and ambient temperature. The prominent factor to remember is that there will be a significant power loss owing to a failure to correspond between the source and the load. In order to get the most power to load from the PV panel, MPPT is implemented in the converter circuit using PWM and a microcontroller. In order to give the most power to load from the source, the solar power system should be designed to its full potential. 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.