Search

The "Daijiworld News" forum, a well-known news website in coastal Karnataka, was the source of the comments for this paper's sentiment analysis study, which was done on about 15,000 reader comments. The comments were scraped using Beautiful Soup, a popular web scraping library and labelled as positive, negative, and neutral. Pre-processing of comments was made using techniques such as stop word removal, tokenization, stemming, lemmatization, and lowercase conversion. Logistic regression, support vector machine (SVM), naive Bayes, random forest, K-nearest neighbors (KNN), AdaBoost, gradient boosting and neural networks was used for classification. Performance metrics including accuracy, precision, recall, and F1 score were evaluated. Logistic regression achieved the highest precision (0.75), recall (0.74), accuracy (0.74), and F1 score (0.74), followed closely by the neural network classifier with a precision of 0.670, recall of 0.670, accuracy of 0.670, and F1 score of 0.669. The study demonstrates the effectiveness of logistic regression and neural networks in sentiment analysis of news forum comments, giving insightful information to grasp public opinion and improving user engagement. The findings contribute to the field of sentiment analysis, emphasising the significance of web scraping and pre-processing techniques in enhancing sentiment classification accuracy. The results serve as a reference for researchers and practitioners, assisting in the selection of appropriate classifiers for sentiment analysis in similar contexts. The study encourages further exploration of advanced techniques to enhance sentiment classification accuracy in regional news forums, paving the way for future research in sentiment analysis. [ABSTRACT FROM AUTHOR]

Efficient inventory management is essential for businesses to optimize their supply chain, reduce costs, and enhance operational efficiency. At the same time due to manual scanning at the checkout processes, customers have to wait for a long time in the queue. As a solution to these different problems, we have proposed a system using IoT. This research paper presents a novel approach to inventory management by integrating two cutting-edge technologies: Load Cell Sensor HX711 and Smart Cart System using Radio Frequency Identification (RFID) technology. By combining the capabilities of these technologies, the paper aims to automate inventory tracking, monitoring, and replenishment, providing real-time data for accurate inventory control and decision-making. The Smart Cart System enhances the customer satisfaction while the Smart Shelves increase efficiency of inventory management by automating the data capture process. It provides convenience by eliminating the need for manual scanning or checkout processes, and generating bill on webpage allowing customers to simply place items in the cart and proceed with their shopping. Also we present the machine learning aspect of our project aimed at revolutionizing the shopping experience using RFID technology and load cells. We explore the implementation of three prominent algorithms - Linear Regression, Random Forest, and XG Boost - to predict future sales. [ABSTRACT FROM AUTHOR]

The temperature profile is an important parameter of the atmospheric thermal state in atmospheric monitoring and weather forecasting. The hyperspectral infrared sounder of a geostationary satellite provides abundant spectral information and can retrieve the temperature profile. Based on the mediumwave channel data (independent variable and model input data) of FY-4A/GIIRS (geosynchronous interferometric infrared sounder) and ERA5 reanalysis data (dependent variable and model output data), the atmospheric temperature profile is retrieved by generalized ensemble learning. Firstly, the feature variables of the model are constructed. Because there are many GIIRS channels, a two-step feature selection method is adopted: step 1--establish a blacklist of GIIRS channels; step 2--select feature variables by using the method of importance permutation. Secondly, they are integrated based on optimizing and adjusting the hyperparameters of three basic machine learning models (Random Forest, XGBoost and LightGBM). Generalized ensemble learning nonlinear convex optimization is used to optimize the weight of each basic model. Finally, based on high-frequency GIIRS observations of Typhoon Lekima and Typhoon Higos, testing and method evaluation of the temperature profile retrievals are carried out. The results show that LightGBM achieves the best retrieval result among the three basic models, followed by Random Forest and finally XGBoost. The root-mean-square error of the whole temperature profile in the training dataset of generalized ensemble learning is less than 0.3 K, while that of the testing dataset is less than 1.4 K, and that between 150 hPa and 925 hPa is less than 1 K. The retrieval results correlate well with the radiosonde temperature profile. The performance of generalized ensemble learning is better than the performances of the three basic models, but it depends on the retrieval results of LightGBM. In the Lekima experimental case, compared to other channels selected for temperature retrieval models, the importance of mediumwave channels 9 and 307 of GIIRS ranks first and second, respectively. The method in this paper provides a new solution and technical support for retrieving atmospheric parameters from hyperspectral and other satellite data. [ABSTRACT FROM AUTHOR]

An aggressive driving state impacts drivers' decisions, which could potentially lead to accidents. Real-time recognition of driving state is particularly important for improving road safety. However, the majority of modeling in existing studies relies on a single algorithm, which may lead to unreliable predictions. This paper proposes a stacking ensemble aggressive driving state recognition model using electroencephalography (EEG), which is able to combine different heterogeneous classification algorithms. Five types of classification algorithms and their variants are tested and compared to identify suitable base classifiers. All of these classifiers are optimized by Bayesian optimizer before the comparison. Three stacking ensemble recognition models using different meta-classifiers (i.e., logistic regression, random forest, and AdaBoost) and an equal-weight voting ensemble recognition model are established. The aforementioned recognition models are evaluated by using a dataset collected from a car-following simulated driving experiment. Fast Fourier transformation (FFT) and wavelet packet transformation (WPT) are adopted to extract features from raw EEG data. The results suggest that the stacking ensemble recognition models outperform the best single (i.e., support vector machine) model; the random Forest stacking recognition model achieves the best performance and the accuracy is increased from 81.21% to 84.23% using FFT features and from 86.45% to 87.38% using WPT features. [ABSTRACT FROM AUTHOR]

With the goal of "carbon peak and carbon neutrality," the need for environmentally friendly gases to replace SF6, a high greenhouse effect gas, is urgent. C4F7N, as an environmentally friendly gas with the greatest potential to replace SF6 as an arc extinguishing medium in circuit breakers, can be mixed with CO2 to greatly improve the shortcomings of its high liquefaction temperature, and the calculation of the physical parameters of the mixed gas plasma is a prerequisite for the computational simulation of the arc process in the opening of circuit breakers. Because solving the physical parameters is expensive, based on the system of differential equations, this paper adopts several machine learning algorithms by mining the relationship between the data using the known physical parameter data to predict the results of the physical parameters to be solved under certain conditions, which greatly reduces the cost of computation. The machine learning algorithms used in this paper are K-nearest-neighbor regression, decision tree, random forest, support vector machine, and gradient boosting regression, of which for the support vector machine, hyperparameters find it difficult to determine the problem of optimization using the gray wolf algorithm. The prediction results of several algorithms show that they are more accurate and that the problem can be solved better by using the method of machine learning. Finally, the comparison results show that the support vector machine exhibits better performance in most cases and that the gray wolf algorithm can make the results of the support vector machine more accurate. [ABSTRACT FROM AUTHOR]

Diabetes is a long lasting disease which affects people worldwide. Many people had diabetes and this number would be increasing in the future. Early detection of diabetes is necessary. Also, diabetes affects other organs and becomes a risk to other organs like heart, lungs, kidneys, skin,eyes, liver. And early detection of the impact of diabetes on other organs is necessary to take proper early precautions and treatment. To identify the early risk, Machine learning can be helpful. Using algorithms of machine learning, it can provide accurate detection and timely diagnosis . This paper contains early detection of diabetes and impact on heart where Random forest is used after model selection to detect the probability of heart disease using UCI heart disease and for diabetes detection PIMA Indian dataset is used. And using this simple analysis it can be found that the diabetic person’s organs are affected or not. [ABSTRACT FROM AUTHOR]

This paper aims to build a machine learning and a neural network model to detect the probability of money laundering in banks. The paper's data came from a simulation of actual transactions flagged for money laundering in Middle Eastern banks. The main findings highlight that criminal networks mainly use the integration stage to integrate money into the financial system. Fraudsters prefer to launder funds in the early hours, morning followed by the business day's afternoon time intervals. Additionally, the Naïve Bayes and Random Forest classifiers were identified as the two best-performing models to predict bank money laundering transactions. [ABSTRACT FROM AUTHOR]

Despite the availability of pipeline bending strain detection technologies based on inertial measurement unit, there is a lack of intelligent and efficient methods for accurately identifying pipeline features by bending strain. Therefore, this paper proposes a novel method for identifying features in natural gas pipelines based on shapelet and blending fusion model. Specifically, the shape features of the bending strain data are extracted and transformed by shapelet. Then a blending fusion model with SVM, Decision Tree and Gradient Boosting as base learners and Random Forest as meta-learner is constructed. Finally, the extracted features are fed into the blending fusion model for pipeline feature recognition. The model is trained with bending strain data obtained from a real natural gas pipeline, the results indicate that the recognition accuracy of the proposed method is 97.17%. Compared with other models, the superiority of the proposed model is verified, and it is proved that the proposed method has better accuracy than the existing models (over 1.3%). Overall, the method proposed in this paper can be effectively combined with the in-line inspection system to provide a reference for pipeline companies to carry out pipeline integrity management. [ABSTRACT FROM AUTHOR]

Organizations have witnessed a significant transformation in the realm of data storage and processing owing to the advent of cloud computing. Nonetheless, with its advantages, cloud computing has also brought forth new security concerns that need to be addressed. In this regard, the use of Intrusion Detection Systems (IDSs) has become indispensable for identifying and preventing a wide range of attacks that may occur in cloud computing environments, thereby ensuring data security. In modern years, machine learning (ML) algorithms have emerged as a promising approach for IDSs, as they can analyze huge amounts of data and identify patterns that may not be detectable by traditional rule-based IDSs. This review paper presents a comprehensive analysis of ML-based IDSs & Tradition-based IDSs for intrusion detection in cloud computing environments. The literature review covers various Traditional & ML algorithms used for intrusion detection, including decision trees, Neural Networks (NN), Support Vector machines (SVM), random forests, and k-nearest neighbours. The performance evaluation metrics used in this review paper include accuracy and false positive rate. These metrics are generally used to evaluate the performance of IDS and ML algorithms. The results of the analysis indicate that ML-based IDSs outperform traditional IDSs in terms of accuracy and false positive rate. However, ML-based IDSs may also have limitations, such as a high rate of false negatives and the usefulness of huge amounts of training data. Overall, the analysis suggests that ML-based IDSs have the potential to improve the usefulness of intrusion detection in cloud computing environments, but further research is needed to address the limitations of these systems. [ABSTRACT FROM AUTHOR]