Search

To develop an automated, reliable and effective system that detects fake news articles, with the goal of reducing the spread of misinformation and promoting the dissemination of accurate and trustworthy information because the rapid spread of fake innovative news has dturn out to be a foremost concern in modern years. Materials and Methods: The effectiveness of two methods Random Forest and K Nearest Neighbor (KNN) are compared in predicting fake news. The evaluation was carried out using a Github dataset of 2000 newscast informations labeled as either counterfeit or unaffected, with performance metrics such as accuracy used to compare the two algorithms. The model size of the group is 10. Results and Discussions: The result shows that KNN outperformed Random Forest (RF) in terms of all the performance metrics, suggesting that KNN is a more effective method for detecting fake news. The significance value for this study is p=0.001 which is less than 0.05. Hence, there is a statistically significant difference between the two groups. Conclusion: The results suggest that KNN with 81.20% accuracy is a more effective algorithm for fake news detection. This study provides valuable insights into the effectiveness of Machine Learning (ML) algorithms in detecting fake-news. [ABSTRACT FROM AUTHOR]

Online purchasing services are quickly displacing traditional or physical retailers. Online shopping websites have seen a considerable increase in customer confidence over time. On the one hand, the proliferation of these websites has encouraged fierce competition, which is excellent for customers because it leads to better and more affordable items. This makes online purchasing a fascinating subject for academic study. Retailers are observing a rise in online transactions from their clients as a result of how simple it is to use E-commerce platforms to make purchases. Predictive analytics can be applied to analyze these interactions and provide intricate behavioral patterns that assist businesses in better comprehending the needs of their clients. Online trust and previous online purchase experiences, along with elements like impulse purchase orientation, brand orientation, and quality orientation, all influence customer online purchase intention and shopping orientation. The objective of our project is to build a prediction model that will help in the increment of the profitability of a marketing campaign for a hypothetical corporation. By applying various preprocessing techniques on data and multiple feature engineering methods, along with four machine learning models we have tried to achieve our goal. The final model should allow the company to focus its advertising on customers who are most likely to respond to the campaign while excluding non-respondents. As a result, four different learning classifiers K-Nearest Neighbors Algorithm, Support Vector Machine Algorithm, Logistic Regression, and Random Forest were tested and optimized, and we have achieved the best classification performance using Random Forest. [ABSTRACT FROM AUTHOR]

The trend of social media interaction, texts, and hidden identity communication has increased drastically over recent years. With the evolution of various methods for convenient techniques of messaging, social media has evolved at the top most position, for almost anyone in the modern world. A major drawback to this "Anonymous Identity" a communication, is that, it becomes very difficult to identify the actual reality of the person communicating from the other ends. This project is purposefully made to implement Natural Language Processing (NLP), Machine Learning (ML), Random Forest Algorithm, K-Nearest Neighbors (KNN) understand behavior traits and search techniques, interests, literal processing of the person, to identify his/her personality, by the input from the most popularly usage platform Twitter. [ABSTRACT FROM AUTHOR]

Book suggestions may be used by users to explore and search for books on the internet. Given a vast number of items and descriptions that correlate to the user's requirements, our recommendation system will assist the user in picking the book that best matches the description. The following criteria impact recommendation algorithms: rating, reviews, description, and author. The effectiveness of Book Recommendation Systems is greatly dependent on the classifier utilized. As a result, developing an accurate classifier is critical for improving the performance of recommendation systems. Decision Tree Classifiers stand out among many supervised learning approaches and algorithms due to their high accuracy, fast classification speed, strong learning ability, and ease of design. The framework for a Decision Tree-Based recommendation system is proposed in this study. Among the other significant supervised learning techniques and algorithms are Naïve Bayes, Random Forest, Logistic Regression, and K-Nearest Neighbor. [ABSTRACT FROM AUTHOR]

The continuous of purchasing online products, For the buy of any product, we must see the reviews of that product after that we purchase the products. Review places the main role in online purchasing the product. Mainly positive reviews make customers buy a product online, similarly, the negative reviews make confusion about whether the customer buys a product or not. The product seller wrote a fake review to satisfy the customer. Thus, identifying fake reviews is a faithful and ongoing research area. Identifying a fake review depends on key features of the review but also it depends on behave of the reviews. This paper proposes a Machine learning accession to the identification of fake reviews. In addition to the features eradication process of the reviews. This paper appeal to some features engineering techniques to essence various behave of the reviews. This paper equates the performance of some trail runs on a real dataset of online shopping reviews with and without features estimations of users' behavior. In both cases, we compare the performance of some classifiers: K-Nearest Neighbor (KNN), Navy Bayes, Logistic Regression, and Random Forest algorithm. [ABSTRACT FROM AUTHOR]

Heart disease is one of the main causes of death worldwide. Machine learning has been discovered tobe useful in creating predictions from massive amounts of data. We've also seen machine learning techniques used in recentadvances in a variety of fields such as medical, finance and even retail. In this research, we used a few traditional ML techniques which is K-Nearest Neighbors (K-NN), Support Vector Machine (SVM), Naïve Bayes, Decision Tree, Random Forest, Logistic Regression, and a Hybrid ML technique that combines Random Forest, SVM and K-NN. We achieved a great performance level with 63.33% accuracy rate of using the hybrid ML model in predicting heart disease. Before applying machine learning techniques, we used feature selection including BORUTA and RFE to identify the Top 10 variables from the dataset to compare with non-using feature selection to build an effective predictive model. Other thanthat, several performance metrics are used to evaluate the results. [ABSTRACT FROM AUTHOR]

In this work, objects from Sloan Digital Sky Survey Data Release 17 (SDSS DR17) [1] were classified as dwarfs or giants using learning methods. The classification was created by supervised learning. Machine learning algorithms like K Nearest Neighbors, Naive Bayes Classifier, Support Vector Classification and Random Forest were developed in addition to Decision Trees. The algorithm that performed the worst was naive Bayes. Random Forest, which performed the best, was able to successfully classify the cases in the dataset that were marked as stars. [ABSTRACT FROM AUTHOR]

The cardiac attack is considered to have high mortality due to lifestyle habits and dietary regimes. Prediction of heart disease based on medical history of patients is critical. As privacy of sensitive data is also of concern, digital fingerprint collected from vast amount of data generated by the healthcare sector is made easier with the help of Machine Learning (ML). With advances in technology application of internet of things (IoT) in healthcare is also gaining popularity. Application of algorithms to predict heart attack is also evident. In the current work, we propose to predict cardiac arrest using medical history for the patients data and select the features best suitable for based on a comparative analysis among the ML models. Different feature selection metrics and classification models are used for prediction. ML algorithms like Regression, Decision tree, Random Forest, Support vector machine, Naive Bayes, and K-nearest neighbor have been used. Our results show the importance of medical history in prediction of cardiac disease and diagnosis. [ABSTRACT FROM AUTHOR]

One of the most common and practical tools for predicting and diagnosing cardiac problems is the electrocardiogram (ECG) signal. It is an advanced medical device fundamental that monitors track of cardiac excitability, transmission, and recovery. It Provides essential information on the functional features of the heart and cardiovascular system. In this work, a technique for classifying an ECG image of a patient case into four classes is proposed. This class includes (MI, Arrhythmia, History MI, and Normal case). There are three steps to the proposed system: The procedure includes "image pre-processing, feature extraction, and classification". In order to prepare and enhance the input images, certain image preprocessing techniques are utilized in the early step. Color images are converted to binary images. Utilizing local binary pattern modulation, features are extracted from the images in the second step (LBP). The extracted features are sent into a 1D ML algorithm as input in the third step. Three ML algorithms are utilized. For ECG classification, these algorithms are "K-Nearest Neighbors (KNN), Random Forest (RF), and Logistic Regression (LR)." Algorithms achieved successful results during the ECG classification process, The findings show that the suggested technique is excellent in RF the range of this classification precision is 93.66 percent, and very good in K-NN the range of this classification precision is 80.53 percent, and good degree in LR the range of this classification precision is 70.87 percent. As evidenced by the low number of fault alarms and high accuracy rate, the suggested system performs exceptionally well in classifying ECG images. [ABSTRACT FROM AUTHOR]

Human activity recognition (HAR) is one of the common classification problems that has a lot of interest. In HAR, different techniques are explored to identity the physical activity of humans, given data coming from various sensors. This paper explored the use of different supervised learning techniques and algorithms such as Gaussian Bayes Theory, Logistics Regression, SVM Classifier, Quadratic Discriminant Analysis, k-Nearest Neighbor and random Forest Classifier. Accuracy measurement of how each algorithm successfully predict the correct classification is done with the Random Forest Classifier giving the highest user accuracy of 87% without cross validation and 84.9% with 5-fold cross validation. For class accuracy, Random Forest Classifier gives an accuracy of 90.5% without cross validation while k-nearest neighbor gives an accuracy of 89.2% with k=3 and 5-fold cross validation. [ABSTRACT FROM AUTHOR]

Heart Disease, Diabetes and lung cancer are complex diseases and globally many people are suffering from these diseases. On time and efficient identification of these diseases, play a vital role in health care monitoring systems. In this article, we proposed an effective and precise framework to determine these and the framework depends on Machine Learning algorithms. The system is developed based on classification algorithms and includes Support vector machine, Logistic regression, K-nearest neighbor, Decision tree, Random Forest and XGBoost Classifiers. Firstly, Kaggle dataset is considered for testing the data. Later pre-processing of data is done. Then the greatest piece of any data analysis project is ensuring that the information is effectively organized and fixing it when it isn't. The initial segment of this interaction is managing Missing Data. After dealing with the missing data, we are ready to start formatting the data set for making classification tree. Formatting data is splitting data into dependent and independent variables. Finally, after formatting the data for classification tree, the data is split into training and testing sets. The experiments results show that the classifier Random Forest tree achieved high accuracy as compared to other classifiers. Additionally, the proposed system can be easily implemented in healthcare for identifying heart disease, diabetes and lung cancer. [ABSTRACT FROM AUTHOR]

Occupancy detection is critical during building operations, particularly in energy efficiency, space utility, heating, ventilation control, and air conditioning (HVAC) to optimize user comfort. However, applying occupancy sensors to detect occupants can violate a person's privacy, and the resulting data could be more precise. Therefore, this study aims to detect occupancy through environmental sensors by utilizing the accuracy of machine learning-based classification results. The ultimate goal of this study is to detect whether there are people in a room based on the benchmarks studied. Benchmarks for classifying whether people are in a room include CO2, humidity, lighting, temperature, and occupancy. This paper presents a method for comparing and evaluating a set of different machine learning techniques based on a given performance measure (for example, precision, accuracy, f1 score, and algorithm recall) and then using the algorithm with the best performance. This study uses five machine learning classification algorithms: K-Nearest Neighbors, Support Vector Machine, Decision Tree, Naive Bayes and Random Forest, and the method used to evaluate the results is cross-validation. Experiments were conducted using occupancy detection datasets from the UCI Machine Learning Repository. After performing the experiment, the predictive results of the five algorithms are high. That is, each has an accuracy value of more than 90%. Support Vector Machine has the highest accuracy value compared to other algorithms, namely 98.5%, and Decision Tree has the lowest accuracy value, namely 91.9%. The results show that selecting the right features and the suitable classification model can significantly impact prediction accuracy. [ABSTRACT FROM AUTHOR]

The IOT, in general, refers to devices being connected to a network. As a result, in order to protect network devices from security breaches, data must be secured. Because of the potential severity of the consequences of an IOT (IoT) failure, further investigation into IoT security vulnerabilities is warranted. The primary goal of IoT security is to protect user-privacy and confidentiality, as well as the architecture, communications, and technology that makes up an IoT ecosystem, as well as to ensure that the ecosphere services become inevitable. Since simulators, designers, and computation and analytic tools have become more widely available, research in IoT security has lately gained a lot of traction. An overview of the present level of IoT-security research, the essential tools, and IoT model and simulators are the core contribution of this study. Machine learning technologies will be used to identify attack threats in IoT security. Decision tree, Support Vector Machine (SVM), Bayes Algorithm, K-Nearest Neighbor (KNN), and Random Forest (RF) are the five most commonly used methods in IOT. [ABSTRACT FROM AUTHOR]

This project aims to detect online payment fraud using machine learning algorithms. Fraud is an illegal criminal activity carried out for monetary or personal gain. Detection of such fraud requires a dataset comprising details about past fraudulent transactions for training, testing, and pattern detection to anticipate any fraud. For this task, authors are using a dataset from Kaggle and have been implementing four algorithms viz. logistic regression, decision trees, random forests, and k-nearest neighbor. However, the data must be cleaned and pre-formatted before training and testing. Therefore, authors perform data pre-processing which removes attributes that do not help to identify fraudulent transactions, handle missing data and outliers, match or encode categorical data, and more. This allows the implementation of a training model to predict fraudulent or non-fraudulent transactions. Finally, the accuracy of all four models is compared and it is noticed that all 4 models are performing quite well for fraud detection. [ABSTRACT FROM AUTHOR]

The analysis of firewall logs is one of the most significant practices considered while monitoring network traffic to assess their impact. The log records of the Turkish Firat University's firewall device were analyzed using K-Nearest Neighbor (KNN), Random Forest (RF), and Deep Neural Network (DNN) classifiers. A comparison was conducted to measure the performance of the classifier in terms of accuracy, recall, precision, and F1 score. 65,532 records were examined using 12 attributes, where the action was identified as a label of these attributes because it handles the packets based on their features either allowing them to pass, blocking them, blocking their activity, or blocking the request itself. The result of this analysis indicated that the best algorithm that selects the best features according to appropriate action is Random Forest. [ABSTRACT FROM AUTHOR]

Crime has become a big and dominating problem in our society so preventing it has become an important task. Currently, the number of crimes that have occurred is increasing and in various forms, this has led to demands for the relevant institutions to develop more efficient methods of taking preventive actions. In line with the current situation, traditional crime prevention methods are considered less effective and produce less optimal results. Every day several crimes occur, therefore technology is needed for the prevention of crime cases more quickly. Maintenance of crime data and data analysis processes can help predict and solve crimes that will occur in the future. The existence of these problems makes the researcher conduct research on how to solve crime cases more easily. Like, if we can find a way to predict crime with the data we already have. This study aims to compare the accuracy of several supervised learning algorithm techniques in predicting crime data in Chicago. 4 different type of algorithm, K-Nearest Neighbor (KNN) classification, Logistic Regression, Decision Trees and Random Forest is used for comparing the different level of accuracy in predicting the type of crime. As the result successfully proved that the KNN, Decision tree and Random Forest have better accuracy than Logistic Regression because the accuracy result are more than 95%. In the last stage, data visualization will be carried out in the form of graphical representations of various insights obtained during the data analysis process. [ABSTRACT FROM AUTHOR]

In India, climate change has a huge impact on the agricultural sector. It directly affects crop yield, nutritional quality and productivity. In this paper, a model is devised to correctly predict the crop yield only from past data including factors like rainfall so that appropriate measures can be taken in the future. Machine Learning is one such tool which can be used to solve this issue. Random Forest Algorithm, XGBoost and K-Nearest Neighbor (KNN) Algorithm are the machine learning algorithms used for crop yield prediction in this paper and the one with higher accuracy can be used for future work. Further discussion includes how much impact each factor related to crop yield has on the predicted value so an ordinary farmer can plant and take care of his crops according to it and use this model for his daily work. In doing this, modernization of the agricultural sector is made possible and a boost is given to crop yield in India. [ABSTRACT FROM AUTHOR]

The aim of the study is to investigate the possibility of applying machine learning methods for saccades and fixation detection from eye tracking signals. There are two widely used approaches to solving this task: velocity-based saccade identification (I-VT) and dispersion-based saccade identification (I-DT). Both require setting an optimal threshold which directly influences detection quality. Another important problem comes from the necessity of noise handling if eye tracking signal has a high noise level. Using I-VT and I-DT algorithms in this case will result in misclassification. Previous researchers have obtained promising results using Random Forest algorithm for saccade and fixation detection. Our aim with the article was to compare Random Forest with other binary classification algorithms such as Logistic Regression and K-Nearest Neighbors. We've also compared mentioned algorithms with velocity-based approach. The results show that Random Forest has the best classification metrics. The findings suggest that this machine learning approach could also be useful for eye movement event detection and proves the importance of further research in this field [ABSTRACT FROM AUTHOR]

Due to the environmental condition people face various health hazards and it affects their living habits. So predicting the disease at earlier stage becomes important. As a specific symptom can be responsible for various diseases, it is time consuming to predict the perfect one for the doctors. Data mining plays a significant role in this most challenging task i.e. the accurate prediction of diseases. Due to increase in data growth in every year in healthcare field the precise analysis of data helps patients to take early care. Using the disease data, hidden pattern information was found by data mining in the enormous amount of medical data. Here in this study based on the symptoms general disease prediction techniques have been proposed. For the disease prediction, K-Nearest Neighbor (KNN), decision tree and random forest machine learning algorithm were used. [ABSTRACT FROM AUTHOR]

Quick disease detection carries a great significance to the medical community. In this pandemic period, when people are falling prey to the deadly virus, it is clearly assume how much the frontline workers are risking their lives. In this period the technology driven laboratories plays an important role. These laboratories detect the dieses symptoms in a short span of time. However, affordability and reach off such medical facilities cannot reach to every citizen of a nation. Therefore, to make the things happening in a affordable way, we proposes an intelligent model for predicting human diseases such as covid19, diabetics, breast cancer, cardio diseases, kidney diseases, and liver diseases by using machine learning techniques. This model is an artificial intelligence based model, where various machine learning algorithms such as k-nearest neighbors, logistic regression, random forest, decision tree, support vector machine, Naive Bayes, and boosting techniques are used. [ABSTRACT FROM AUTHOR]

Though activity identification has already been studied for such a long period of time, the bulk of research and implementations have concerned with the physical activity recognition. Despite the fact that effective activity detection is necessary in a wide range of applications, research upon those activities has gotten less attention. One reason for this discrepancy is the scarcity of datasets to evaluate and compare different approaches. We developed the Open Laboratory Nursing Activity Identification Challenge, that focused on the detection of complex nursing duties, to stimulate study in such scenarios. The nursing field is one of the most likely to gain from activity recognition due to a scarcity of datasets. The competition was using the CARE COM Nurse Care Activity Dataset, which consisted of seven activities completed by eight persons in a controlled environment using accelerometer, motion tracking, and an indoor location sensor. In this article, the results of the competition are summarized. In this study, we employed widely used approaches such as Decision Based Tree Classification (DT), Random Forest Classification (RF), Support Vector Machine (SVM), and K-Nearest Neighbors (K-NN). The accelerometer data was 71.2 percent accurate, the Meditag data was 81.1 percent accurate, and the Mobility Capture (Mocap) data was 99.99 percent accurate when we employed those approaches. [ABSTRACT FROM AUTHOR]

Security plans for Internet of Things devices are scarce due to their numerous benefits, including the wide range of controller plans. They rely only on unavoidable cross-sorting IoT malware to control disparate concerns. Ingenuity research relies on machine learning algorithms to keep its office alive with malware expressions. Machine learning approaches such as Naive Bayes, K-Nearest Neighbor (KNN), Decision Trees, and Random Forests are utilized in this work to examine malware protection. This study compared all four Machine learning techniques based on 7k malware samples with 38 labels with two subsets. The training phase uses the first subset with 5k data, whereas the testing phase uses 2k data. According to an examination of different show quantifications, the Random-Forest model beats other models in numerous heuristics. [ABSTRACT FROM AUTHOR]

Despite considerable success in discovering the knowledge, conventional machine learning algorithms may defeat to achieve satisfying performances during the transaction with imbalanced, complex, noise, and high dimensional data. In this context, it is substantial to think about efficiently building an adequate knowledge and mining model. Ensemble learning aims to consolidate the classical machine learning (ML) algorithms, data modeling, and data mining into a unified framework. Text categorization is a critical application that uses the unified ensemble learning framework to detect a new article's class. This paper develops a two-layer stacking ensemble model containing different ML algorithms. Since, stacking model consist of stacked layers and each layer built with multiple ML algorithms we constructed the first layer of our stacking model with three ML algorithms (Multinominal Naïve Bayes (MNB, logistic regression (LR), and k-Nearest Neighbor (k-NN)) classifiers, while the second layer applies a random forest classification algorithm. The proposed stacking ensemble model is compared with the classical ML algorithms MNB, LR, and k-NN) in accuracy and error measure. The result shows that using the stacking model outperforms better than MNB and k-NN algorithms with Accuracy reached 89.72% and 89.75 %, respectively. While using LR, the Accuracy equals 91.5%, which is closed to the result of the proposed model, which equals 91.66%. [ABSTRACT FROM AUTHOR]

Over few decades, there are quite a number of clinical guidelines are provided to healthcare workers regarding treatment suggested for homogenous patients, facing similar medical symptoms. Certainly, such recommendations may differ among guidelines, which results in tough scenario for clinicians to make a better decision. The scope of the work is to recognize homogeneous patients having similar medical signs and symptoms, demographics and health data. Distinguishing proof of comparative patient cases is helpful to work on clinical results. In view of the proposed model utilizing Machine Learning (ML) Algorithms like K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Naive Bayes, Random Forest, and expert opinion is utilized as a base for infection analysis and treatment idea. Better accuracy of 98% is achieved by SVM. It helps in the treatment or medication proposal of another patient, expectation of clinical outcomes, clinical help desk, research and in numerous other cases. [ABSTRACT FROM AUTHOR]