Fake news is misleading information which can often damage individual’s reputation and can change person’s opinion. Recently, the rise of social media platforms gave a boost to circulation of fake news across the globe. Our paper presents a machine learning model using voting ensemble with equal and different weights of each base model. We used 4 different classification algorithms as base learners- 1- Logistic Regression, 2-Multinomial Naive Bayes, 3- Passive Aggressive Classification Algorithm, 4- Random forest algorithm. All algorithms chosen works best with the long textual in stream of data. We compared accuracy of each base model. After assigning different weights to each base model, voting ensemble gave an accuracy of 93.5
In recent years, there has been a surge in interest in a few features of high-tech commercial drones. Small drones have a higher potential for being used for illegal operations due to their ability to evade ground security while transporting payloads. Security breaches like this may only be avoided with drone tracking and monitoring. Recognizing drones in surveillance footage can be challenging due to the similarity between small drones and birds, especially against complex backdrops. Keeping an eye out for drones and other flying objects manually is a time-consuming and difficult task. Therefore, it is necessary to employ a mechanical means of telling drones apart from birds. In this research, we create a system for drone identification using focus measure operators (FMOs). On every video frame, the five FMO parameters are calculated. Drone identification begins with a feature ranking to determine which features are most important and then continues with a classification using a random forest (RF) classifier. The Workshop on Small-Drone Surveillance, Detection, and Counteraction Techniques (WOSDETC), with funding from the Safe Shore Consortium, provides the data used to assess the suggested method’s efficacy in the Drone-vs-Bird Detection Challenge at IEEE AVSS2021. The suggested method presents to identify drones with drone present (DP) vs neither drone nor bird present (NDNBP) (two class), DP vs both bird and drone present (BBDP) vs NDNBP (three class), DP vs BP vs BBDP vs NDNBP (four class) with average acc 94.15
As the popularity of Transformers in computer vision rises, it is likely that Transformer-based models will become the standard for many types of vision applications. The Transformer-based model’s unmatched scalability and vast quantity of trained data make it tough for literatures with fewer data then less computational skills to apply it. Because of its patch-based representation, Transformer-based models have recently been shown to perform very well, and a study presented ConvMixer to demonstrate this. ConvMixer excels in picture classification, but its wasteful isotropic design makes it unfit for other vision tasks. In this research, we offer HEConvMixer, a patch-based representation network that is both hierarchical and data-efficient. In contrast to the original Transformer-based models, our network features two down sample layers and some simple convolutional blocks in place of the Transformer blocks. Our network was trained from scratch on modest datasets using a single GPU. Our HEConvMixer achieves 98.12
The rampant increase in the spread of misinformation around the globe on the social media in the wake of the pandemic international conflicts has urged the innovation in machine learning paradigms of the fake news detection to tackle the threat. Given that the task belongs to the text classification under the subfield of Natural Language Processing, the benchmark defying performance of the transformers-based models such as BERT inspired to use the pre-trained model for text classification using two approaches, one using the BERT for feature extraction and then classifying the text using standard classifiers, a methodology which has not been widely researched, while other involving fine-tuning the BERT for text classification. The various classifier models achieved remarkable accuracy up to 98
In cloud cryptosystems, a diversity of advancement can be experienced to make strong both the authentication and data storage (compression) in open cloud environment. This paper expands the standing effort of authentication with compression schemes. The earlier drawbacks include normal schemes for client’s metrics, direct authentication verification with risks of information outflow to attackers, uploading of replica content and slow speeding of the network. To overcome these existing flaws, this effort delivers a framework deploying a trusted server. When the clients access to open cloud, they share the identities and then accelerated to open cloud with authentication keyword. For compression, the modified deflate algorithm is incorporated for better performance. This strategy helps in generating better results for authentication and data storage milestones. The main objective of the proposed work is to secure the data in the cloud from the unauthorized users and improve the storage capacity in cloud. The proposed framework is sub divided into three stages: Requisition phase, authentication phase and storage phase. The primary stage involves the requisition phase which uses the basis login and password requisition model. The secondary stage executes authentication and authorization. In the final stage the data is compressed and stored in the network using modified deflate data compression algorithm.
Modern machine learning (ML) and deep learning (DL) techniques are combined with sensory techniques like electronic nose and tongue, spectrum imaging, and color recognition to conduct an in-depth analysis of different types of brewed tea. This research proposes a novel technique for comprehensively evaluating tea, including its classification, nutritional profiling, and health benefits, that challenges conventional assessment methods. The complexity arises from the great range of teas available and their potential effects on health. In this study, we combine state-of-the-art ML and DL methods with sensory instruments for an in-depth analysis. The paper begins with a thorough literature review that reveals the technological development of tea analysis, the benefits and drawbacks of current approaches, and opportunities for future research and development. Powerful ML and DL algorithms are deployed and supplemented with sensory data, all on top of a comprehensive collection of high-definition tea photos. The accuracy shown in classifying and identifying teas is a testament to the efficacy of the proposed method. Further, these algorithms accurately identify nutritional components and evaluate health benefits, bringing a new angle to the study of tea. Accurate data collection, model selection, and sensory data integration are highlighted, and a delicate balance between interpretability, robustness, and computing efficiency is emphasized. The research concludes by highlighting the promise of this multi-modal approach in practical applications within the tea business, despite the difficulties inherent in real-time implementations. This ground-breaking work integrates technology and the food sciences, suggesting potential uses for other consumables.
Finding and labelling expressions of opinion or emotion in written content is the primary focus of sentiment analysis. Now more than ever, people turn to online communities to share their perspectives and vent their emotions. As a result, there is a mountain of data being produced each day that may be successfully mined for insights. Conducting sentiment analysis on this kind of data can help generate a holistic perspective on certain items. Sentiment research on Twitter can be difficult because of the widespread use of slang and misspellings. The rapid influx of fresh words also makes it harder to analyse and compute the sentiment than it would be with more conventional sentiment analysis methods. The maximum number of characters allowed in a tweet is 140. Therefore, another challenge is overcoming the limitations of short communications to learn crucial details. Sentiment analysis of tweets can greatly benefit from knowledge-based techniques and machine learning. As individuals adapt to new ways of interacting on social media platforms like Snapchat, Instagram, Twitter, etc., the amount of data they generate increases at an exponential rate. There are literally billions of fresh pieces of content uploaded every day, including text, music, and video. This is due to the fact that numerous people frequent the website in question. These individuals wish to express their views on whatever subject they feel fits. These entries are meant to express the thoughts of a single individual on a particular subject. The goal of this research is to analyze these posts and identify the feelings that motivated their creation. We’ve settled on Twitter as the venue for this effort. The changes made to this social networking site are referred to as “tweets.” In this research, we look at how Twitter users feel about specific businesses. Critical feedback on the company’s products from people all around the world would be offered by generating a basic sentiment score and then classifying it as positive or negative.
This study explores the transformative potential of AI-driven healthcare solutions in underserved areas with limited access to medical facilities and resources. The aim is to investigate how AI technologies can bridge the healthcare gap and improve healthcare delivery for populations facing geographical, infrastructural, and socioeconomic barriers. The study explores various AI applications, including telemedicine, diagnostics, predictive analytics, and remote patient monitoring, while addressing challenges such as connectivity, privacy, and cultural considerations. Through an examination of case studies and current initiatives, this study highlights the opportunities and limitations of implementing AI in remote healthcare settings and underscores the importance of ethical, cost-effective, and sustainable approaches.
This paper presents an innovative approach to address traffic congestion and safety challenges in smart cities by leveraging Artificial Intelligence (AI)-driven Vehicular Ad-Hoc Networks (VANETs) within IoT-enabled transportation systems. The integration of AI algorithms, such as machine learning and deep learning, enables seamless communication among connected vehicles and IoT infrastructure. Real-time data analysis facilitates effective traffic flow control, congestion detection, and accident prediction at the same instance security and privacy concerns are addressed through robust solutions. The current work showcases simulations and case studies, highlighting significant improvements in traffic efficiency, reduced travel time, and enhanced transportation safety. The study emphasizes the transformative potential of AI-driven VANETs in creating intelligent transportation systems for future smart cities, fostering more sustainable and liveable urban environments.
Recent years have seen a surge in interest in crop recommendation systems that consider the market and the weather to help farmers choose the right crops. The best crop can be predicted using machine learning approaches and algorithms based on market data, demand, and supply, and this survey article gives an overview of the available research in this area. The survey emphasises how crop recommendation systems use well-known techniques like Random Forests, Support Vector Machines, and Artificial Neural Networks. It covers its use in emerging market analysis, supply-demand information, and climatic factors to give farmers precise advice. Also mentioned are the drawbacks of the research papers under review, such as the lack of data, the narrow geographic reach, and the development of new technologies. The survey’s results highlight how machine-learning approaches can boost agricultural output and profitability while taking market dynamics and climatic conditions into account. Future research could address the survey’s limitations to create more reliable and useful crop recommendation systems that are adapted to various agricultural environments.
In the modern world, chronic kidney disease (CKD) can have a devastating impact on human survival. Accurate and timely detection of CKD is of great importance for the prevention and treatment of renal failure. Non-invasive techniques such as Machine Learning models (ML) provide a high degree of reliability and efficiency in distinguishing healthy individuals from those suffering from CKD. The goal of this research is to develop a unified method for predicting CKD by applying the classification algorithms of ML to the UCI repository dataset and the medical records of affected individuals. To build the prediction system, the authors used all the basic preprocessing methods from ML and a hybrid extraction technique to reduce dimensionality. Analysis of the results with two different datasets showed that the highest accuracy of 95.83
Concerns regarding the health and well-being of the old have been raised globally due to the elderly population’s rapid rise. In order to overcome these difficulties, this study introduces an Internet of Things (IoT)-based wearable health monitoring system for senior care. The suggested system combines a variety of sensors with wireless communication capabilities to continuously monitor the health and physical activities of elderly people. The central element for data collection, processing, and transmission is the Arduino platform, which is renowned for its simplicity and adaptability. In addition to an accelerometer for tracking movement and fall detection, the wearable gadget has sensors for assessing vital signs like heart rate, blood pressure, body temperature, and oxygen saturation. The Arduino board processes the data locally, and it is then wirelessly transferred to a centralized monitoring system using Wi-Fi or Bluetooth Low Energy (BLE) communication. Caretakers or medical experts can access the central monitoring system, which offers real-time visualization and analysis of the gathered data. It makes it possible to discover aberrant health situations early, to act quickly, and to perform remote monitoring. Additionally, the system may produce warnings and notifications in the event of crises, ensuring quick access to medical care. The suggested wearable health monitoring system built on Arduino has a number of benefits, including affordability, mobility, and simplicity of use. It gives elderly people the ability to preserve their independence while offering a safety net through ongoing health monitoring. Additionally, the system enables healthcare professionals to give individualized and timely interventions based on real-time health data and simplifies remote caregiving.
In healthcare, IoT-based applications are growing day by day to perform predictions regarding chronic diseases using machine learning. There is a widespread consensus that melanoma, or cancer of the skin, is one of the worst illnesses in the world. A precise classification of skin lesions in their early stages might likely assist in the process of therapeutic decision-making, therefore improving the probability of a cure before cancer develops. The likelihood of developing cancer of the skin is highest in those parts of the body that are often exposed to the damaging effects of direct sunlight. In men, these body parts include the head, face, lips, and ears; in females, the chest, arms, and hands; and in both sexes, the legs. On the other hand, it is also possible for it to develop on sections of your body that are seldom exposed to air and light, such as your hands, feet, and other spots on your body. Deep learning has been considered a subset of machine learning that is frequently used to develop detection and classification mechanisms. In the current study, the deep learning methodology is being considered as a potential method for detecting skin cancer in a shorter amount of time and with more precision. Compression operations have been performed to increase performance, and hybrid deep learning models have been used to improve accuracy measures such as recall, precision, and f1-score. In the IoT environment, individuals can benefit from continuous and real-time skin health monitoring. As ML algorithms continue to evolve and gain access to extensive datasets, their role in skin cancer detection and recognition within IoT holds immense promise for enhancing healthcare and prevention strategies.
The paper presents a cutting-edge real-time feedback detection system utilizing advanced face recognition technology. It analyses and interprets the user’s facial expressions to provide instantaneous feedback on their emotional state. By employing state-of-the-art technologies and algorithms in face recognition and emotion detection, the system delivers accurate and prompt feedback to enhance user engagement and performance. The study involves designing an intuitive and user-friendly interface that seamlessly integrates with the models. Rigorous system testing and user experience tests have been conducted to ensure the system’s functionality and effectiveness. The outcomes includes the delivery of a fully functional system capable of accurately detecting real-time feedback, thereby facilitating improved user engagement and performance. Also, the challenges of obtaining accurate and comprehensive feedback for business meetings, seminars, and movies are examined in this work. The proposed model provides the ability to capture emotional responses to events and content as they occur and can provide a more objective and standardized approach to feedback gathering.
The world is under the turbulence of the transformation from the age-old traditional healthcare systems to contemporary, patient centric and clinicians need based operations. The introduction of Artificial Intelligence in the healthcare industry though besets with a bunch of demerits deserves special mention with respect to its over brewing merits. In contrast with the global standard, India, a developing country has obvious financial and infrastructure specific bottlenecks in bringing the success of the mass implementation of Artificial intelligence in the healthcare operations. However, the genuine and continuous efforts are made in streamlining the critical and stereotyped operations for the benefits of the medical service seeker and also for the competitive survival of medical service provider. The present study focused on the historical development of Artificial Intelligence in healthcare, the reason of its gradual popularity, the application of the tool, some of the notable used cases where the Artificial Intelligence gained its momentum and a host of pros and cons in dealing with it. The study also featured the futuristic intensity of its application of Artificial Intelligence in the healthcare units in India to ease out the pain of availing the emergent medical services without the typical intervention of the medical experts and on the contrary also the administering the hassle-free diagnostic procedures with transparency and smart approach.
Artificial Intelligence (AI) in education is revolutionising the way that learning experiences are personalised and instructional environments are defined in contemporary pedagogy. This study examines the various ways that artificial intelligence (AI) is changing education, with a focus on how these innovations have the potential to transform conventional teaching strategies and improve learning experiences for students of all ages. (Mendoza, Rose Marie N. and Dayao, Edna F. 2021 [20]). AI-driven learning environments provide adaptable and customised learning experiences catered to individual needs as traditional classroom structures change. AI algorithms may determine students’ learning preferences, skills, and weaknesses through the examination of data-driven insights. This information enables teachers to tailor their pedagogical strategies. AI-powered educational systems can also give prompt feedback, which encourages self-evaluation and progress. This study looks at several AI-related applications in education, such as virtual mentors, intelligent tutoring systems, and intelligent content recommendations. With the use of these technologies, students may participate in realistic simulations, gain access to a multitude of materials, and get help instantly - all of which enhance the quality and inclusivity of the learning environment (Dabbagh, N. Bannan-Ritland, B. 2005 [7]). AI also makes it easier to automate administrative duties, freeing up teachers’ time for instructional design and one-on-one student interactions. AI-powered analytics also support curriculum creation and programme assessment, helping educational institutions stay flexible and adaptable to the ever-changing demands of their students. Even if AI has a lot to offer education, privacy and ethical issues need to be taken care of. This essay examines algorithmic biases, the ethical ramifications of data collecting, and the possibility that artificial intelligence will worsen educational disparities. It highlights how important it is to implement AI in educational settings in a transparent and responsible manner. This study concludes by highlighting the ways that AI is improving educational settings and personalisation, which in turn is empowering education. It emphasises how AI has the ability to raise teaching standards, boost student performance, and create a more inclusive and egalitarian learning environment. To optimise the beneficial effects on education, it also emphasises how crucial ethical issues and responsible AI application are.
Malaria is a life-threatening disease that affects millions of people worldwide, particularly in developing countries. Early and accurate detection of malaria is crucial for effective treatment and control of the disease. In recent years, deep learning techniques have shown promising results in various medical imaging tasks, including malaria detection and diagnosis. This paper presents a comprehensive review of deep learning applications for malaria detection and diagnosis. It covers the different stages of the malaria diagnosis pipeline, including image acquisition, pre-processing, parasite detection, and classification. The review discusses various deep learning architectures employed in malaria detection, such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), and their variants. It also highlights the challenges and limitations of existing approaches and identifies potential areas for future research. The findings of this review demonstrate the potential of deep learning techniques in improving the accuracy and efficiency of malaria detection and diagnosis, ultimately contributing to the efforts in eradicating this global health burden.
The most popular technologies nowadays are wireless sensor networks, which have advantages like low cost, small size, and mobility. But the networks like Wireless Sensor Network (WSNs) are resource constrained with respect to energy utilization that comprises tiny solitary sensor nodes with limited bandwidth. Furthermore, some of the other main issues to pay attention are: accumulating sensed information from the network, transferring the data to the base station while focusing on network coverage, lifetime, and power conservation. More researchers have recently become interested in using machine learning techniques in wireless networks to solve these problems that arise in the specified network. This paper reviews the existing approaches of machine learning that have already proposed. Furthermore, detailed analysis of exiting work also discussed in the form of table that serve as a reference for anyone interested in learning more about designing suitable Machine Learning solutions for the wireless sensor netwok’s applications.
Orthopedic implant identification is the most important step before performing any revision surgery. Automated identification of implants plays a huge role in preoperative planning. All the existing studies on automated identification of implants through artificial intelligence use X-ray images of implants as both training and testing data. The proposed novel and unique work identify the make and model of knee implants from 2D templates and their corresponding preprocessed images in contrast to the use of X-rays in training. The proposed work uses X-ray images of implants only for testing. The experimental approach with different deep learning models provides classification accuracy of 50.83
Emails are utilised in practically all spheres of today's society, from the professional world to the academic sphere. Ham and spam are the two subcategories that may be found within emails. Email spam, also known as junk email or unwelcome email, is a sort of email that may be used to cause harm to any user by wasting his or her time, using an excessive amount of computing resources, and stealing important information. The proportion of unsolicited emails is rising at an alarming rate day by day. Predicting the value of a company's stock is difficult for academics, investors, and analysts. The majority of people are interested in learning about stock prices in order to enhance their own finances. Long-Short-Term Memory (LSTM) is the abbreviation for the time series notation. In today's market, a stock trading system needs to adhere to this paradigm and combine KNN and LSTM in order to achieve higher levels of accuracy in its models. The majority of people in today's world improve their financial situations by trading on the stock market. When this doesn't work, individuals’ resort to criminal behaviour. The two equities are compared using this procedure. In order to solve the problems that the pure KNN method was having with distance metrics, the suggested model uses an optimised version of the KNN technique. The fact that the majority of test data in the present KNN is focused on focal points has no impact on the stock prediction because all of the qualities are connected. The Programme places a greater emphasis on data points that are favorable rather than ones that are centered. The KNN distance probability is determined by an optimization procedure. The present iteration of the KNN model demands a significant amount of memory in addition to other resources so that it can compute the distance between each data point and the test point. By removing duplicates, the procedure eliminates the need to double-check the records. The procedure is sped up by reducing the number of iterations. The effectiveness of the model is demonstrated by comparisons with standard classifiers.
