Search

Clinical laboratory testing is a crucial part of delivering high-quality health care. Lab tests are ordered by a doctor or another clinician to determine, control, or demonstrate a patient's condition. The process starts with the preparation of a pattern of the patient's blood, tissue, or other organic material, which is then sent to a lab to be uniquely identified and checked to make sure it is appropriate for the test. Some exams are graded by hand, but the majority are completed with the aid of technically advanced equipment. The electrocardiogram (ECG)is a crucial tool for identifying cardiovascular issues. The huge number of ECG records were formerly recorded on paper. Examining paper ECG recordings manually can be challenging and time-consuming. The Machine learning algorithms, Support vector machine and other algorithms in particular show promise for systems that have abilities to detect cardiovascular illnesses where the effectiveness differs in terms of accuracy, recall and precision. The proposed research provides a thorough analysis of several data models for predicting patient illness for clinical diagnosis in cardiology. [ABSTRACT FROM AUTHOR]

Various machine learning algorithms are developed with an aim to create precise and trustworthy models and extract knowledge from data sources. Deep expertise in the field of machine learning is required for the challenging task of choosing the right algorithms for a specific dataset. There is no single algorithm that outperforms all others across all applications and different datasets. The difficulty of choosing an appropriate algorithm for a specific task in specific domain is related to the properties of the dataset. Properties of the dataset are measured through meta-features. Meta-features describe task and can provide explanation how one machine learning approach outperforms other algorithms on a given dataset. Learning about the effectiveness of learning algorithms, or meta-learning was developed to deal with this issue. Focus is required because previous research papers have not successfully identified meta-features in particular domains. In this research, we have evaluated various meta-feature characterization methodologies and have concentrated on basic meta-features. Business domain data is in the focus of this paper. We computed basic (general) meta-features and illustrated several use cases for their applications. [ABSTRACT FROM AUTHOR]

Face Emotion Detection Music Player is an innovative idea that combines computer vision and machine learning techniques to create a unique and interactive music player. This Research Paper studies the uses of Convolutional Neural Networks (CNN) to detect and analyze facial emotions in real-time and dynamically create playlists based on detected emotions. The music player is designed to record facial expressions using a webcam or other suitable camera. To reliably identify facial expressions including happiness, sorrow, rage, surprise, and more, a CNN model is trained on a sizable collection of face images. The trained model demonstrated an astounding 74.46% accuracy in identifying emotions when combined with music player software running on a PC or other suitable device. The CNN model analyses the photos to determine the user's emotional state as they interact with the music player in real-time, as captured by the camera. Based on the detected emotions, the player automatically selects and plays songs from a predefined music collection that matches the user's emotional state. For example, if the user looks happy, the player can play happy and energetic songs, while if the user looks angry, it can play more calming and soothing music. Face Detection Music Player offers a personalized and dynamic music listening experience as the music selection is constantly updated according to the user's emotions. This research paper illustrates how computer vision and machine learning techniques may be used to produce interactive programs that can instantly adjust to the moods and preferences of their users. It has uses in user experience research, entertainment, and mental health, among others. [ABSTRACT FROM AUTHOR]

There are various models out there, both theoretical and experimental, which are used to predict the fatigue crack growth in metals. Machine learning based computational models have been recently developed to predict such crack growth rate in a more comprehensive manner. This paper presents a supervised machine learning model with a better accuracy than the existing machine learning models. The experimental data obtained from the fatigue test carried out on CT specimen of material SA516Gr70 is used to test and train the model. The machine learning model XGBoost, which is implemented in this paper, is not only more accurate but also less likely to over fit when exposed to unseen data. Three alternative algorithms are investigated and by the process of elimination, best one is chosen. The paper shows the results received from some of those models, which were important in order to select the best one. The metrics used to filter and tune the model's performance are R2 score and mean squared error. The procedure for finding the optimal model and its results have been discussed. [ABSTRACT FROM AUTHOR]

The detection of chest diseases from X-ray images is a crucial aspect of medical diagnostics, playing a vital role in the early identification and treatment of respiratory conditions. This review paper gives a general overview of how machine learning algorithms are used to identify chest diseases from X-ray pictures. It emphasizes the importance of accurate and timely diagnosis of chest diseases, while acknowledging the challenges faced by conventional diagnostic methods. The review highlights the potential of machine learning as a promising approach to enhance diagnostic accuracy and efficiency in this field. It explores the utilization of both handcrafted features and deep learning-based approaches for extracting informative features from chest X-rays. Furthermore, it discusses the adoption of various machine learning algorithms, including support vector machines, random forests, and convolutional neural networks, for effective detection of chest diseases. The review paper also recognizes the significance of training and validation strategies in ensuring the robust development of models. Additionally, it addresses the potential impact of the proposed methodology on chest disease detection and patient outcomes. The review paper discusses challenges and limitations associated with this approach, such as data availability, interpretability, and ethical considerations. In summary, this extensive analysis strives to make a valuable contribution to the field by offering insights into cutting-edge machine learning techniques, recognizing existing challenges, and proposing potential avenues for future research. [ABSTRACT FROM AUTHOR]

This state-of-the-art review paper provides a comprehensive overview of meta-learning techniques for learning to learn in machine learning. Meta-learning called "learning to learn," is an approach that enables models to learn from new tasks learning quickly and efficiently, based on prior knowledge and experience. The paper discusses various meta-learning techniques, such as model-based meta-learning, metric-based meta-learning, optimization-based meta-learning, and memory-based meta-learning. The review also covers the applications of meta-learning in different domains, including natural language processing, computer vision, and reinforcement learning. Additionally, the paper discusses the benchmark datasets, challenges along with the future directions of meta-learning research, highlighting the need for developing more efficient and scalable meta-learning algorithms. Moreover, the proposed Meta-Learn framework incorporates a hierarchical structure to facilitate knowledge transfer across multiple levels of abstraction, and it also includes a memory component that enables the system to store and reuse previous experiences. Overall, this review aims to provide a comprehensive overview of Meta-Learning techniques for learning to learn in Machine Learning and proposes a novel Meta-Learn framework that could potentially enhance the performance and scalability of Meta-Learning algorithms. [ABSTRACT FROM AUTHOR]

The COVID-19 outbreak impacted drastically to education and most of educational institutions started preferring online education for students. However, after the settlement of the pandemic there is uncertainty among people about whether they should prefer online education for furthermore or start in offline mode to make it more interactive, so this paper is about an analysis of people's sentiments and emotions through Tweets about COVID-19 Education. This paper aims to study the reaction of people around the world toward online education during COVID-19. This study is conducted on the basis of the responses of students, teachers, parents, college professors, etc. We started with labeling the data into three sentiments namely positive, neutral, and negative and for validation then we used Machine learning (ML) classifiers namely, Logistic regression, Decision tree, Random Forest, Multilayer Perceptron (MLP), Naïve Bayes, Support vector machine (SVM), K-nearest neighbors (KNN), and XG-Boost. Then we performed emotion detection by considering 5 emotions namely happy, surprise, sad, fear, and angry and for validation we used ML classifiers. After applying all these ML approaches, the XG Boost ML classifier achieved the highest accuracy of 94% in classifying the tweets as positive, neutral, or negative, and 96% accuracy in classifying the tweets as happy, surprised, sad, fearful, or angry. [ABSTRACT FROM AUTHOR]

One of the main forces advancing Industry 4.0, also known as the fourth industrial revolution, is machine learning (ML). This paper examines the machine learning (ML) models application for biogas production prediction in anaerobic digestion (AD). This study's primary objective is to determine which ML techniques and models are utilised in the AD process. In addition, this study identifies the type of ML techniques, input and output parameters, and software used. Researchers have widely employed a couple of ML models in biogas production. After reviewing the 15 most recent papers, it was discovered that the Artificial Neural Network (ANN) and Adaptive Network-Based Fuzzy Inference System (ANFIS) are the most commonly used types of ML. The most commonly used operating parameters in predicting biogas production were reaction time, temperature, pH, total solids (TS), volatile solids, volatile fatty acids (VFAs), and fixed solids. In conclusion, the review discusses the challenges and prospects of using machine learning in the AD process and provides recommendations for future implementation. [ABSTRACT FROM AUTHOR]

The sine and cosine algorithm has become a widely researched swarm optimization method in recent years due to its simplicity and effectiveness. Based on the advantages, the study in this paper delves deeper into the key parameters that influence the performance of the algorithm, and has implemented modifications such as integrating the reverse learning algorithm and adding elite opposition solution to create the modified Sine and Cosine Algorithm (the modified SCA). Furthermore, by combining the fuzzy k-nearest neighbor method with the modified SCA, the study simulates numeric datasets with two or multiple classes, and analyzes the results. The accuracy rate (ACC) achieved by the modified SCA FKNN in this paper is compared to other models, with data comparison results and tables presented for each. The modified SCA FKNN proposed in this paper has obvious advantages on accuracy rate(ACC). [ABSTRACT FROM AUTHOR]

In the machine learning era, aspect-based sentiment analysis is a very popular idea that is still in the research realm. The main objective of this paper is to derive new pattern-based rules for Aspect Extraction. This focuses on the Explicit aspects where the aspect is precisely defined, unlike the previously researched papers on Aspect Extraction, this model does not limit itself to the noun/noun phrases but also focuses on the other words like verbs in a review. Newly extracted pattern rules for aspect extraction would be proved beneficial in terms of sentiment analysis of online customers. [ABSTRACT FROM AUTHOR]

Today, the world is exposed to many widespread health crises. One of these crises is the emergence and spread of the covid-19 virus, where any person becomes infected with this disease through the rapid transmission of infection between people, either by palpation or by inhaling droplets of carriers of this virus. Hence, there is a need to build a tracking system to respond quickly to problems. This paper proposes using one of the unsupervised machine learning algorithms, the K-Means Clustering Data, to determine the areas with the most significant percentage of infections by dividing them into infected and uninfected human groups. Where the infected persons are divided into three high groups (C1), medium (C2), and low (C3), and follow the path of the infected persons using Google Maps. The contribution to this paper is to support appropriate decision-making by public health authorities by predicting endemic areas and how to address problems that occur as they can use this system to reduce other infectious diseases. [ABSTRACT FROM AUTHOR]

Health and medicine has gained a lot of importance in today's digital world, where evolving technology is being used to fight against almost all the known diseases thoroughly. But, according to reports, more than two lakh people in China and one lakh in United States of America are dead every year due to the mistakes made while prescribing errors. This paper aims to propose a system which takes as input the symptoms from the patient to predict the disease, which is followed by recommending the correct medicine. This system consists of database system module where disease dataset contains attributes like "DiseaseID", "Disease", "Symptoms" and "count of occurrences" and drug dataset has additional attributes such as "drug" and "drug_rating". In the data preparation module, a new data frame is created where each disease with its associated symptom is given "1" and "0" if it is not the symptom related. After all these steps of preprocessing are performed on the raw dataset, the dataset is now ready to be acted upon by a machine learning algorithm. Models like Decision Tree, Random Forest and Naive Bayes were applied on the training-dataset. Entire knowledge database was fitted in models and it was seen if not much dataset is used Random Forest And Naïve Bayes classifiers showed 84% accuracy whereas Decision Tree gives 92%. But Random Forest also gives accuracy similar to Decision Tree for large dataset. Hence for disease prediction we use Decision Tree. The recommendation is done on the basis of the rating of the drug. After this, only the drug with the highest rating is included in a newly formed dataset and the rest of the drugs are ignored. The system is implemented for user access by making use of Tkinter a python GUI Library for constructing the desktop end user interface. This paper deals with the design and implementation of a system which performs the function of prediction of diseases and the recommendation of medicines, which adds to the capabilities of the present systems in health infrastructure. [ABSTRACT FROM AUTHOR]

This paper discusses a keen way to manage and distribute water from a water tank or reservoir to consumers with minimal wastage. The aim of this paper is to discuss about the implementation of an IoT (Internet of Things) based water distribution system that will diminish the water wastage by consumers. The setup can be extended from a small society to the entire urban infrastructure. The proposed system consists of a waterflow sensor, actuator, microcontroller (ESP8266), and a web-application which can remotely control, cut, monitor, and analyze the water usage. Administrators will have access to restrict the water flow and upon consumers' requests ancillary water can be supplied. Also, this system is equipped with machine learning techniques for predicting the usage and learns the flow patterns of water among different user spaces, which will be useful to estimate the water requirements. Also, classify consumers based on their usage. The application is equipped with a reward system where consumers are rewarded with ancillary water when their consumption is minimal. [ABSTRACT FROM AUTHOR]

Twitter is known to be one of the familiar social networking platform these days, among many others, with a lot of user engagement. This microblogging site encourages social interactions, allowing users to stay up to date on the latest news and events and share them with others in real time. Tweets are limited to 280 characters and is allowed to include links to related websites and tools. With a platform having such wide reach, it is prone to be targeted negatively and spams are one way to do it. Spammers use this platform to display malicious content that is inappropriate and harmful to users worldwide. Machine Learning uses various approaches that can be used to detect spam and overcome it. However, with the advent of recent technologies it has been observed that the properties of tweets vary overtime making it difficult to detect spam leading to the "Twitter Spam Drift" problem. This paper reviews the papers published since 2018 that have focused on the spam drift problem and gives a comparative analysis of the different algorithms that are utilized on the various data sets to tackle such a problem. [ABSTRACT FROM AUTHOR]

The aim of this paper is to carry out work those results in a good friend to farmers in every way. The work primarily recommends a high-yielding crop to the farmer based on regional soil data and the farmer's requirements. The job entails analysing a database of different types of soil from a specific area. To improve the analysis' efficiency, a machine learning method called linear regression is used. The study also implements disease detection in crop leaves using image processing techniques and provides remedy/solutions to farmers with diseased leaves, which promises to be a preventative measure in the future. The study carried out has an effect on various cultivation methods for various crops, assisting farmers as a "SAHA PATI" for this work. This service assists farmers in determining the best method of cultivation based on climatic and geographical factors. The final section of the Paper introduces a programme that provides current market prices for various crops by connecting to a government website that is updated on a regular basis. This service assists farmers in avoiding being duped by business intermediaries. The overall work done with the agriculture-based component turns out to be the best application for the current situation. [ABSTRACT FROM AUTHOR]

This paper proposed to analyze various DDOS ATTACK protection mechanisms for cloud environment. DDOS ATTACK is very big threat for networks and communication. Now a day every sensitive information stored in cloud environment and other types of dual services such as PAAS, SAAS, IAAS were also attacked by DDOS ATTACKS. So cloud environment must be protected from these types of DDOS ATTACKS. In this paper we provide complete outlook for various DDOS protection, such as machine learning methods with various other methods. Different types of methods tested with test data downloaded from various authentic sources in this paper shows structured ATTACK and prevention. Architecture such as test website ATTACKED by various DDOS ATTACK tools and prevention method executed by separate module or integrated with ASP.NET website. [ABSTRACT FROM AUTHOR]

This paper presents the detailed overview on the artifacts related to Computed Tomography (CT) specifically Cardiac CT (CCT), their classification and various reduction techniques. In CCT, the motion artifacts occurs mainly due to the pumping of the heart which has to be reduced during and post CT reconstruction. The main drawback of CT machine is that it emits high radiation during scanning. Therefore, Low Dose CT (LDCT) and Limited Angle CT (LACT) have been proposed by various researchers to overcome the problem of emission of high radiation during scanning which provides better image quality with minimum radiation. However, the artifacts generated during scanning and in image reconstruction phase also affects the quality of the image which makes it a challenging task to medical experts during their interpretation. In view of that, the extensive research have been carried out by various researchers to reduce the noises or artifacts using traditional and machine learning techniques in recent past. Finally, the paper is concluded by presenting the current state of art deep learning approaches to enhance the performance of the existing methods. [ABSTRACT FROM AUTHOR]

The first, and at the same time one of the most important tasks in building a predictive system is the task of analyzing the original data set. The next task is to process the data according to the purpose and objectives of the predictive system. These two tasks are often not given enough attention and, as a consequence, the developer does not get the desired result as a result of building the system. This paper discusses methods and algorithms for data analysis and processing that can help in building a predictive system. The data set from the machine learning competition "ASHRAE - Great Energy Predictor III" is used as input data. The analysis phase focuses on constructing graphs and drawing conclusions from them. The paper shows how to use graphical information to find anomalies in the data and eliminate them. Examples of graphs that provide information useful for building machine learning models are given. The data processing phase describes the transformations performed on the data. It considers the transformations that lead to positive results as well as those that lead to negative results. [ABSTRACT FROM AUTHOR]

Machine learning is an artificial intelligence (AI) technique for automation and prediction that is widely utilized today. This paper looks at the fledgling usage of machine learning in the legal field. The first portion focuses on the fundamentals of machine learning. The relationship between machine learning and law is discussed in the next section. Many parts of the law can be interpreted in terms of data, and machine learning works by recognizing patterns in data. Machine learning technologies can be used to learn about data-driven aspects of the law. After that, this paper looks at how machine learning is used in law while also recognizing the technology's strengths and limitations. It finishes by examining some current social debates around the use of machine learning in legal contexts. [ABSTRACT FROM AUTHOR]

In this paper, the Intensity of the Characteristic Q-learning (InCh Q-learning) algorithm is proposed. It allows for finding the shortest and at the same time the safest escape route. Data on the intensity and spread of a fire occurring in a virtual electrical substation are used. Matrices are entered with the intensity values of each considered fire characteristic. As the fire spreads in space, zones of intensity are formed. Rules are introduced that take into account intensity zones for finding an escape route. The learning algorithm finds the shortest path for which the intensity of each of the dangerous features is the least. A threshold of the intensity of each character is introduced at which a person can pass and evacuate successfully. The use of priorities for the dangerous features allows one to choose a path with a greater intensity of the safer features and at the same time with a lower intensity of the more dangerous ones. The optimal escape route found is used to build a decision tree to find the location of an injured user as quickly as possible. [ABSTRACT FROM AUTHOR]

The literature review found that only a few studies in Indonesia have been applied to modeling peatlands fire occurrences. In our previous studies, various machine learning classification approaches are applied to predict forest fire occurrence in the peatlands area. Our previous empirical study found that the datasets of fire hotspot areas, especially for data from the South Kalimantan area, need to be more balanced. In this paper, we consider ADAptive SYNthetic (ADASYN) method for the sampling approach to balance the datasets. We found that by balancing the data, the performance of the classification method can improve about 2-4%. [ABSTRACT FROM AUTHOR]

It is known that the studies of peatlands fire occurrences in Indonesia are less studied before. In our previous study, the prediction of the peatlands fire occurrence was modeled using various machine learning classification approaches. It is found that using South Kalimantan Province data, in the empirical study we previously found that the datasets are unbalanced, i.e., the occurrence and the nonoccurrence of fire hotspots areas. In the study presented in this paper, to improve the classification performance, we consider Adaptive Neighbor Synthetic Majority Oversampling Technique (ANS-SMOTE) approach to balance the data. Using the considered empirical data, we found that this method did not always gives improvement in the classification results. [ABSTRACT FROM AUTHOR]

Multi-class image classification is one of the tasks of computer vision in which convolutional neural networks are widely used. In this paper, we create a novel ensemble learner by stacking InceptionV3, Xception, MobileNetV2, and ResNet50 convolutional neural network in an attempt to outperform the best performance of each individual models. In an ensemble learning algorithm, by combining multiple different models with different architectures, an error of prediction from a single model can be corrected by other models, thus reducing the likelihood of selecting a wrong prediction. The dataset that we used to do benchmarking is the Intel Image Classification dataset which consists of 6 classes that displays images of fully colored natural scenes. The training set consists of 14034 images, while the test set consists of 3000 images, all having 150x150 as their default size. The training process on all models involves reducing their learning rate when the performance plateaus, and early stopping if there aren't any significant improvements from an epoch to the next. The result is that each individual models have an average of 83% accuracy, yet our ensemble model managed to reach 88% on the test set. [ABSTRACT FROM AUTHOR]

Techniques for detecting oral cancer are a developing area in medicine that includes cancer diagnosis and early detection. For the classification of oral cancer, we have tried to develop machine learning models like logistic regression (LR), naive Bayes (NB), K-Nearest Neighbors (KNN), support vector machine (SVM), decision tree (DT), and random forest (RF), as well as hybrid models like KNN-SVM and SVM-KNN and deep learning models like Xception, VGG, CNN, Inception, and MobileNet. A few researchers have already used the aforementioned models to categorize oral cancer using genetic data. Nevertheless, the database used for this study includes pictures of the lips and tongue that have been divided into malignant and non-cancerous categories. With the assistance of ENT doctors, the photos were categorized after being captured in various ENT hospitals in Ahmedabad. To determine which model is the most effective, the ML, hybrid, and DL models are evaluated based on numerous evaluation criteria. Experimental result shows that the accuracy, precision, recall, and Fl score of classification of the ML, Hybrid, and deep learning are discussed in this research paper. Out of which hybrid model SVM-KNN with PCA has 74%, Gradient boosting with PCA with 73%, and deep learning model VGG has I 00% accuracy. Hence, it may be concluded that Gradient boosting with PCA and advanced deep learning models like VGG, and Xception is the most favorable model for predicting the survival rate of oral cancer patients by classifying them into cancerous or non-cancerous. This research study motivates towards the early prediction comparatively reduces the severity, and economic cost as the treatment for early diagnosis is less. [ABSTRACT FROM AUTHOR]

Joint Time-Frequency Analysis (JTFA) is a well-known concept for studying time-varying signals in engineering and scientific applications. Short-Time Fourier Transform (STFT) is a widely used JTFA technique due to its superior response for real-time applications. However, the apriori's fixed window size has limited time and frequency resolution. There are several advanced techniques proposed in the literature to overcome the limitations of STFT, such as Wigner-Ville Transform (WVT), Continuous Wavelet Transform (CWT), S-Transform (ST), Generalized S-transform (GST), Complex S-Transform (CST), Bi-Gaussian S-Transform (BGST), and Hyperbolic S-Transform (HST). These advanced (JTFA) techniques demonstrated good time and frequency resolution of time-varying signals, resulting in better fault detection and classification in many engineering and scientific fields, especially for condition monitoring. In addition, the performance of these JTFA algorithms under noisy conditions is widely available in the literature. In this paper, the authors investigate the performance of S-Transform-based JTFA techniques for time-varying signals under noisy conditions and highlight their superior performance in selecting a suitable method for specific real-time industrial applications. In this study, several non-stationary test signals proposed in the literature were simulated using Matlab. The joint time-frequency analysis of these unique non-stationary signals was conducted under both normal and noisy conditions for the performance evaluation of various transforms/algorithms mentioned earlier. The results obtained are suitable for developing real-time monitoring and JTFA tools for condition monitoring and diagnostic studies with the help of IoT and Machine Learning for Smart Health Monitoring and Asset Management. [ABSTRACT FROM AUTHOR]

The reality is that the population and number of automobiles on the street are growing day by day. The efficient monitoring of continuously increasing traffic is a critical challenge that traffic operators are facing. Traffic control is generally achieved using stationary video recorders, observers or manual counters. Hence, Unmanned Aerial Vehicles or drones are the high-quality answers to this trouble as they lessen the price of the set-up of numerous cameras and sensors. This paper's principal ideology is split into phases. The first one is the monitoring of traffic using a drone which helps in analyzing the traffic in a particular area and making interpretations. The second is fire detection and extinguisher. Here the drone is used to detect fire with the aid of a fire sensor as it helps the fire-fighters to analyze the affected area and prioritize their work. This project also up-skills in designing a quad-copter. Our motivation for this project lays both personal interests in a better understanding of object detection and academic research. A drone consists of a camera to cover a wider area which helps in traffic analysis. Here, the number of vehicles are estimated. Drones have recently emerged as an outstanding technology in the business market and are geared up with a camera, geo-positioning sensors, and communications hardware that may relay statistics. This project is implemented by using artificial intelligence and the Internet of Things for Fire Detection. To be more specific traffic detection is achieved using Deep Learning algorithm which is the Convolutional Neural Networks Algorithm using python. Fire sensors are mounted on the drone, and it flies on a flight path within the building and sends fire alert data to the system controller upon sensing a fire condition via fire sensor and also puts off the fire using the water pump. For fire detection the sensor used in the drone basically detects IR (Infra-Red) light wavelength that is emitted from the fire flame which helps fire-fighters to quickly assess and monitor dangerous areas which results in more safety and limits the time it takes to scan the whole area. [ABSTRACT FROM AUTHOR]

Bidirectional Encoder Representations from Transformers (BERT) is one method technique in machine learning for solving various natural language processing (NLP) problems, including summarization. One general problem using BERT is getting much time and resources to train the model. This paper aims to implement the small architecture BERT, called Small2BERT, which is expected to summarize the data. Therefore, this study compares pre-trained Small2BERT and Summarization using Word Frequency (SWF) in extracting text summaries. This research uses the dataset of Indian News Summary, which includes news articles from the Hindu, Indian Times, and Guardian from February to August 2017. Small2BERT surpasses SWF in precision, whereas SWF gives a higher score in overall high f1-score. Small2BERT earns a 0.307 f1-score in Rogue-1, compared to 0.35 for SWF. Small2BERT scores 0.19 and 0.304 while SWF achieves 0.304 and 0.31 in f1-score for Rogue-2 and Rogue-L, respectively. [ABSTRACT FROM AUTHOR]

Internet of things is a little world of connected things with the integration of smart devices like sensors, actuators, gateways, gadgets etc. Smart cities, Smart home, Agriculture, Disaster management, Smart Grid, Healthcare system are some of the major applications of the Internet of Things. Machine Learning is another paradigm of the technology revolution and used as a predictive analytics technique. It is employed in about all the areas for the analysis of data. In this paper we proposed a system for the diagnosis and prediction of diseases by unification of Internet of Things and Machine Learning. Data generated by medical sensors, gadgets is fed to semantic engine to make it semantic interoperable and Machine Learning is further employed for predictive analytics. Semantic engine used Resource Description Framework (RDF) to convert data into a common format and data is retrieved using SPARQL queries and GraphDB. [ABSTRACT FROM AUTHOR]

Activity recognition offers a wide range of applications in various industrial processes and healthcare. This work proposes an approach to collect data from a spherical coordinate system using smartphones, then extract the highly efficient features using advanced preprocessing. The paper also proposes an algorithm to recognize activity using various ensemble machine-learning approaches based on extracted features. These approaches are evaluated under various combinations of features to analyze the accuracy, sensitivity, specificity, and training time. Experimental results reveal that weighted KNN performs best among all models by achieving 96.2% accuracy with 12 features. On the other hand, Bagged tree ensemble classifiers perform better than subspace KNN ensemble classifiers with an accuracy of 95.3% using 12 features. [ABSTRACT FROM AUTHOR]

With the continuous increase of internet usage in today's time, everyone is influenced by this source of the power of technology. Due to this, the rise of applications and games is unstoppable. A major percentage of our population uses these applications for multiple purposes. These range from education, communication, news, entertainment, and many more. Out of this, the application that is making sure that the world stays in touch with each other and with current affairs is social media. Social media applications have seen a boom in the last 10 years with the introduction of smartphones and the internet being available at affordable prices. Applications like Twitch and YouTube are some of the best platforms for producing content and expressing their talent as well. It is the goal of every content creator to post the best and most reliable content so that they can gain recognition. It is important to know the methods of achieving popularity easily, which is what this paper proposes to bring to the spotlight. There should be certain parameters based on which the reach of content could be multiplied by a good factor. The proposed research work aims to identify and estimate the reach, popularity, and views of a YouTube video by using certain features using machine learning and AI techniques. A ranking system would also be used keeping the trending videos into consideration. This would eventually help the content creator know how authentic their content is and healthy competition to make better content before uploading the video on the platform will be ensured. [ABSTRACT FROM AUTHOR]

Assessing type of soil and its required nutrients is an important domain in modern agriculture. So moving towards the same vision, this research work addresses this issue of forecasting suitable crops on the basis of environmental factors and its yield based on previous data sets available on the net. The study not only discusses yield and crops, but also stresses on the amounts and types of nutrients present in the soil beforehand by using supervised machine learning algorithms. Among different models used, random forest generates the best performance. Further in the paper we will see how random forest provides us an accuracy of 93% and the least error rate of only 0.3% among all other algorithms using rainfall as a parameter to predict our desired crops. [ABSTRACT FROM AUTHOR]

Women psychiatric health is one of the most neglected factors. Women mental health is very important as mentally healthy woman is the backbone of entire family and society. Even though many papers are published related to mental illness detection and treatment using various methods through survey, questionnaires, public health databases, clinical records, social media usage etc., by applying different Machine Learning (ML) algorithms for prediction, still there is huge room left for identifying and analyzing the mental illness through different means and applying ML for prediction, detection and assessment. The main aim of this research work is to demonstrate that speech is the easiest way to recognize symptoms of mental illness like depression, stress, anxiety, trauma etc. The speech signal carries hidden attributes like intensity, pauses, speech rate which reveal lot of information about the psychological fitness of a woman. The model is deployed on all the kernels of SVM to study and analyze the prediction accuracy of classification using three classification labels. The result obtained is more realistic in assessing the psychiatric fitness with overall accuracy of 90.78 percent. [ABSTRACT FROM AUTHOR]

In today's world, being the easy availability of the internet, almost everyone is connected. Every individual for their daily news relies on online resources, which greatly impacts their quality of life. The development of social media platforms has grown tremendously. However, many of these platforms are spreading fake news that pauses serious concern. It leads to a decrease in income for individuals or companies. Some people often create fake news for gaining fame, for election purposes, and for creating hatred towards other companies to reduce their revenue. All these can have negative consequences on society. Just by the use of a single click, many people forward appealing headlines, photos, etc. to gain the attention of an individual, without even checking whether it is fake or not. It is hard-core truth that many false websites give false news and checking their authenticity is really a challenge. This paper presents a comparative analysis of methods for the user to check the news whether that's real or fake by providing accuracy and recall using machine learning. The models used are naïve Bayes (NB), decision tree classifier (DT), random forest (RF), and logistic regression (LR) for analysis to detect fake news. The result shows that the DT classifier achieved 99.56% accuracy as compared to RF (99.35%), LR (98.91%), and NB (94.89%). The recall also shows better results for DT and RF than LR and NB. Lastly, it discusses some of the security threats and challenges faced while detecting fake news. [ABSTRACT FROM AUTHOR]

A topic that evokes a large base for assorted studies, across numerous countries is soil classification. Recently, the world recorded an exorbitant rise in human population, the denouement is a higher demand for food and shelter. Both these factors directly demand exceptional study in geo technology, if not, a more reliable and faster method to do so. Conventional strategies taken by farmers are insufficient to meet the skyrocketing demands, and as a result, they must impede soil cultivation. This leads to the requirement to remain cognizant of the optimum soil type for a certain crop to maximize agricultural production. Several laboratory and field methods for classifying soil exist, however, these have constraints namely time and labour sine qua non. A need for a software-based soil categorization modus operandi that will assist humankind from a range of engineers to farmers, in reducing the time deviation is sought. This paper discusses apropos software, that allows soil classification to be achieved more efficiently. The functioning includes image processing and Computer vision-based soil classification methodologies, embodying the customary image assessing algorithms based on parameters, namely texture, colour, and particle size. [ABSTRACT FROM AUTHOR]

This research paper explores the application of text classification and natural language processing techniques for enhancing hate speech detection. The study employs machine learning (ML) and deep learning models, including transformer models such as BERT, RoBERTa, and DistilBERT, to improve the accuracy of hate speech classifiers. Through a comprehensive empirical analysis on three diverse datasets (Data-ICWSM, Data-ALW2, Data-OLID), the study demonstrates the effectiveness of these models in accurately identifying hate speech. Compared to traditional baselines, the BERT models exhibit a significant performance boost in macro and weighted F1-scores. Additionally, the study addresses the challenge of imbalanced class distributions in the datasets by employing sampling techniques during training. Overall, the research highlights the potential of transformer models for hate speech detection and provides insights for future exploration in this domain. [ABSTRACT FROM AUTHOR]

Network security has emerged as an essential issue as an outcome of the Internet's broad consumption. The efficiency of anomaly-based detection systems for network intrusions (IDS) as a tool for spotting illicit traffic has risen. IDS's precision and efficiency have been enhanced with the assistance of ML methods. The performance of multiple machine learning (ML) algorithms in anomaly-based intrusion detection is compared in this paper using KDD-CUP-99 dataset. The algorithms considered include Voting, LightGBM, Decision Tree, KNN, Random Forest, AdaBoost, Naive Bayes Model, CatBoost, and Logistic Regression. The study's findings indicate that the Decision Tree, Random Forest, LightGBM, and Voting algorithms did remarkably well and exhibited high rates of accuracy, while Naive Bayes performed inadequately. This study comes to the conclusion that using ML algorithms might greatly improve the precision of knowledge-based network intrusion detection and provide a workable network security solution. [ABSTRACT FROM AUTHOR]

Artificial Intelligence is an incessantly evolving and progressing technology. Machines can exhibit humanoid intelligence, every now and then even better than humans. Its models are being extensively used in today's sphere. It is due to the supremacy of data and artificial intelligence. However, in Machine Learning, the data is amassed through numerous devices, including smart gadgets like smart watches, laptops, mobile phones, etc., and is transported to a centralized server. These devices are also called edge devices. When the personal data of users is transferred to the servers where it is used to train models it creates issues related to data privacy. Federated Learning (FL) is contemporary to technologies like Blockchain, on-device AI, IOT, and Edge computing where private/personal data of the user is not sent on the server, rather training of models is carried out locally, on each device. While the nodes are learning continuous communication between them takes place in Federated Learning. Therefore, high bandwidth connections that can exchange parameters of the machine learning model along with adequate computing power and memory of the local devices are desired. Therefore, one of the major challenges involved in FL includes communication efficiency. This paper presents a survey about FL and discusses the recent developments related to overcoming constraints in communication efficiency in Federated Machine Learning. [ABSTRACT FROM AUTHOR]

The paper analyses the experiences and expectations of the consumer about the product during online purchases, the supply and demand of the online market, and the promises made by the company. Differentiate e-commerce products based on similar features, price, or quality. The tools used on the dataset are Pandas, NumPy, NLTK, and SkLearn. A better product is distinguished based on the ratio of positive to negative edges. Customer satisfaction, sales volume, and rankings were also determined. These factors were the main reason for customer satisfaction. The research looks at customer expectations and experiences with online buying, including product preferences, usability, and product promises. The competitive market environment, which makes it challenging to identify desired items from identical ones in terms of features, price, or quality, is said to be the reason why buyers are cautious about various products. [ABSTRACT FROM AUTHOR]

Run 3 of the Large Hadron Collider (LHC) of the data-taking period poses unprecedented challenges to the computing models used in the high-energy physics experiments of the LHC accelerator. Only in the LHCb experiment, the luminosity has increased by a factor of five. Recent results show that deep learning solutions techniques can significantly improve the performance of the cluster reconstruction in calorimeters when high occupancy is expected. In this paper, we will review selected results of the LHC experiments and, in particular, focus on the investigated convolutional (CNN) and graph neural network (GNN) solutions for planar, LHCb-inspired calorimeters with hybrid granularities. We will also show how the most recent developments in the new experiment-independent simulation framework – Gaussino – can be used to produce training datasets for these models. [ABSTRACT FROM AUTHOR]

With the upcoming large-scale surveys like LSST, we expect to find approximately 105 strong gravitational lenses among data of many orders of magnitude larger. In this scenario, the usage of non-automated techniques is too time-consuming and hence impractical for science. For this reason, machine learning techniques started becoming an alternative to previous methods. In our previous work [1], we have proposed a new machine learning architecture based on the principle of self-attention, trained to find strong gravitational lenses on simulated data from the Bologna Lens Challenge. Self-attention-based models have clear advantages compared to simpler CNNs and highly competing performance in comparison to the current state-of-art CNN models. We apply the proposed model to the Kilo Degree Survey, identifying some new strong lens candidates. However, these have been identified among a plethora of false positives, which made the application of this model not so advantageous. Therefore, throughout this paper, we investigate the pitfalls of this approach, and possible solutions, such as transfer learning, are proposed. [ABSTRACT FROM AUTHOR]

In many scientific fields, Bayesian models are commonly used in recent research. This research presents a new Bayesian model for estimating parameters and forecasting using the Gibbs sampler algorithm. Posterior distributions are generated using the inverse gamma distribution and the multivariate normal distribution as prior distributions. The new method was used to investigate and summaries Bayesian statistics' posterior distribution. The theory and derivation of the posterior distribution are explained in detail in this paper. The proposed approach is applied to three simulation datasets of 100, 300, and 500 sample sizes. Also, the procedure was extended to the real dataset called the rock intensity dataset. The actual dataset is collected from the UCI Machine Learning Repository. The findings were discussed and summarized at the end. All calculations for this research have been done using R software (version 4.2.2). [ABSTRACT FROM AUTHOR]

Phishing attacks have become increasingly common in recent years, with cybercriminals using a variety of techniques to trick unsuspecting victims into providing sensitive information such as login credentials or financial data. One of the most common methods of phishing is through the use of URLs, where attackers create fake web pages that mimic legitimate sites and use them to steal information from users to combat this threat, researchers and security experts have turned to machine learning techniques to develop algorithms that can accurately detect phishing URLs. In this paper, review the current state of the art in URL phishing detection using machine learning, including the various approaches and algorithms that have been developed. [ABSTRACT FROM AUTHOR]

Quantitative structure-activity relationships (QSAR) describe the relationship between quantitative chemical structural properties (molecular descriptors) and biological activity. QSAR assays are increasingly used in drug discovery and development as they can save significant time and human resources. Several parameters affect the predictive performance of QSAR models. On the one hand, various statistical methods can be used to study the linear or nonlinear behavior of a data set. Feature selection approaches, on the other hand, are used to reduce model complexity, limit the risk of overfitting/overtraining, and select the most important descriptors from hundreds of lists. A mathematical model is then used to relate the selected descriptors to the biological activity of the corresponding molecule. A variety of modeling strategies can be used, some of which involve explicit feature selection. QSAR models are useful for developing new compounds with increased potency in the class under consideration. Only connections that are considered interesting are created. Learning algorithms face the challenge of selecting a meaningful subset of features of interest while ignoring the rest of the feature selection problem. This paper studied the comparative analysis of the Chi-square, Mutual Information, Anova F-value, Fisher Score, Permutation Importance, Recursive Feature Elimination, Random Forest, LightGBM and SHAP feature selection methods used in QSAR modeling. The Python code written to get experimental results in this article has been uploaded to Github (https://github.com/kushmuratoff/feature_selection). [ABSTRACT FROM AUTHOR]

Cervical cancer is one of the most known health problems faced by women around the world. Early detection of cervical cancer may reduce the mortality rate. Pap smear images are new techniques used for screening cervical cancer. This paper excludes the nucleus of pap smear images, and the resultant images are classified into seven classes based on the surrounding region nucleus. Automated features are extracted using three pre-trained convolutional neural networks (CNN). The resultant features are twenty-one. The principal component analysis reduces the dimensionality and selects the most significant features into ten features. These features are fed to two types of machine learning algorithms: support vector machine (SVM) classifier and random forest classifier. The support vector machine classifier achieved the highest accuracy for seven classes, reaching 93.1%. This method will help the physicians in the diagnosis of cervical cancer depending on the tissues, not the nucleus. Furthermore, the result can be enhanced using a huge amount of data. [ABSTRACT FROM AUTHOR]

Heart disease forecasting is a key issue in the clinical data analysis field. Artificial Neural Network (ANN) is a machine learning tool that can help doctors diagnose heart diseases more accurately and quickly. However, the design of ANN is complicated because it needs to identify optimal weight values and suitable network structures. This paper aims to optimize the weights of ANN for forecasting the existence of heart disease among humans by using Artificial Bee Colony (ABC) algorithm to train ANN and select optimal network weights. The experimentations are carried out on the UC Irvine Machine Learning Repository (UCI) heart disease datasets and tested by the MATLAB software machine learning tools. The accuracy, specificity, sensitivity (recall), precision, and F-score of the ANN trained by the ABC model (ABC-ANN) and ANN trained by ABC with backpropagation (ABC-BpNN) model are investigated. Between ABC-ANN and ABC-BpNN, it was shown that the ABC-BpNN model has a better predicting ability and could reach considerably higher accuracy with 90% and 90.9%, 88.9%, 90.8% and 90.9% for precision, specificity, recall, and F-score, respectively. The findings showed significant enhancements compared with previous studies that have utilized the same dataset. [ABSTRACT FROM AUTHOR]

The efficiency of learning algorithm can be improved by introducing propositional satisfiability in Discrete Hopfield Neural Network and modeling the neuron structure of the neural network. To improve the flexibility of logic structure and to meet the requirements of all combinatorial problems, a special satisfiability logic structure is introduced in Discrete Hopfield Neural Network. In this paper, Exhaustive Search was harnessed as the learning algorithm to search the neuron fitness, and a constant learning trial was conducted for the sake of implementation. In order to study the learning mechanism on the special satisfiability logic, we focused on the influence of different learning trials on the structure and evaluated the performance of the model at different phases following the evaluation of indicators. We also analyzed the effect of different learning trials on synaptic weights in learning phase and discussed the quantity and quality of global solutions obtained by the model in testing phase. It has been suggested that the learning mechanism was that the higher the trial value, the better the performance in the learning phase and the testing phase. From the perspective of solution diversity, the learning mechanism embodies better solution diversity in fewer trials. [ABSTRACT FROM AUTHOR]

COVID-19 is an infectious disease which is spreading fast and some variants of the coronavirus, such as Delta and Omicron, are spreading more easily between people and this spreading can be alleviated to some extend by vaccination. In this context, the government is demanding for producing the vaccination certificate for the person to be in certain public places. But there can be chances where people do not know their Aadhar number and they do not carry their vaccination certificates especially in rural areas. And also it takes a lot of time in waiting in a long queue for verification purpose. So to tackle all these problems we propose an automated system for identifying vaccinated person without much human interventions. The system uniquely identifies the person through iris recognition using various deep learning models. In this paper we made a comparative analysis of different deep learning and CNN architectures (MobileNet, EfficientNetB3,InceptionV3, DenseNet169) for building a model for person recognition and select the best model with maximum accuracy. Here, we used kaggle dataset (1000 iris images) for testing and training. Also using an iris scanner, the live iris images can be captured and used for model building. Using some deep learning algorithms, the features of iris image of the person is obtained and it is compared with those in Aadhaar database to get the Aadhar number. Using this Aadhar number, the details of the vaccinated person can be obtained from vaccination database and system will identify whether a person is vaccinated or not. By installing our proposed system, vaccinated person can be easily identified without much human interventions, avoiding touch and reducing the long queue for verification, thereby maintaining social distancing. [ABSTRACT FROM AUTHOR]

Since their emergence, high dimensional data have imposed a big challenge to researchers given their complexities. To deal with this problem, using feature selection is the best solution, it plays a very important role in high dimensional data analysis. This process will first allow us to extract the relevant features, remove redundancy and noise, decrease the computation time and improve the classification performances. In the literature, there are several methods for feature selection, of which three categories can be distinguished. In this paper, we will focus on the filter methods for feature selection, this category estimates the quality of the variables individually independently of any learning algorithm based on statistical measures. Among the most popular filters, there is the Relief method created by Kira and Rendell as an instance-based learning-inspired idea, it gives weight to each feature and selects the highest-scored features. In this study, we propose a filter method for feature selection called ReliefTP. ReliefTP is a combination of the Relief concepts and the Technique for Order Preference by Similarity to the Ideal Solution(TOPSIS). Several experiments were carried out on four biomedical datasets the most famous including a comparison with the original Relief. The classification results achieved by the different classification methods confirm the effectiveness of our proposed method. [ABSTRACT FROM AUTHOR]

The prevalence of gaps and outliers within datasets presents substantial challenges, particularly in the realm of time series fore-casting and various other predictive machine learning (ML) tasks. This paper, introduces an effective technique for correcting gaps and outliers in data and validates the approach by applying it to datasets with outlier zones drawn from three diverse contexts. This innovative technique holds promising potential to enhance the performance of machine learning models by treating the data to alleviate the complications posed by these issues and in doing so contributes a valuable tool to the data science toolbox. [ABSTRACT FROM AUTHOR]

India is one of the hugest rice cultivating country in the world. For thousands of years rice is a most major food, consequently every mankind depends on it directly and indirectly. But the Main hindrance in rice production is rice leaf diseases. It greatly reduces the production rate. Rice leaf diseases have specific symptoms that cannot be easily identified by usual human's eye. Using one of the image processing techniques, the rice leaf diseases are detected efficiently. In order to detect the diseased image Median Filter, K-means Clustering, GLCM and SVM classifier algorithms were proposed. Accuracy, Precision and ROC of SVM classifier were calculated as performance metrices. Around 97±2 % accuracy was achieved in this method. This paper discusses the efficient methodology for rice leaf disease detection. By using proposed method diseased leaves are identified at the early stage, which can aid farmers in preventing further damage to their crops. [ABSTRACT FROM AUTHOR]