Outstanding success of CNN image classification affected using it as an instrument for time series classification. Powerful graph clustering methods have capabilities to come across entity relationships. In this study...
详细信息
Traditional risk assessment methods face challenges when dealing with massive data and complex operating environment. Therefore, this paper proposes a hybrid mining algorithm-Integrated Risk Decision Tree (IrdT) which...
详细信息
ISBN:
(数字)9798331519032
ISBN:
(纸本)9798331519049
Traditional risk assessment methods face challenges when dealing with massive data and complex operating environment. Therefore, this paper proposes a hybrid mining algorithm-Integrated Risk Decision Tree (IrdT) which integrates decision tree and random forest. IrdT algorithm combines the intuitive interpretation of decision tree and the integrated learning advantages of random forest, and improves the accuracy and stability of risk assessment through the training and integration of multiple decision tree sub-models. The experimental results show that the accuracy of IrdT algorithm on the test set reaches 90.40%, which is significantly higher than that of logistic regression model (76.80%) and support vector machine (81.20%). In addition, IrdT algorithm shows a fast response speed when dealing with real-time data, which verifies its effectiveness in complex and multi-dimensional risk assessment tasks. The risk assessment model based on IrdT algorithm constructed in this paper not only improves the accuracy of risk assessment of key operating vehicles, but also provides detailed risk classification and quantitative assessment methods, which provides strong support for the safety management of transportation industry. The research results show that big datamining algorithm has important application value in risk assessment of key operating vehicles.
Artificial Intelligence (AI) is an umbrella term for systems that can act in cognitive processes in a human-like and human-enhancing manner, e.g., in learning, problem solving, and patternrecognition. According to mo...
详细信息
ISBN:
(纸本)9781914587238
Artificial Intelligence (AI) is an umbrella term for systems that can act in cognitive processes in a human-like and human-enhancing manner, e.g., in learning, problem solving, and patternrecognition. According to models of technology adoption, several factors influence the actual implementation of a new system within an organization and in an individual's professional practice. These factors include e.g. job relevance, demonstrable results, individual experience with technology, and voluntariness to adopt the new system. This research studies employees' views and expectations of AI applicability and its impact to teachership within a Finnish higher education institution (HEI). Survey data was collected from different schools and units from all hierarchical layers of the HEI, a University of Applied Sciences. Views on AI were assessed in relation to the core tenets of a teacher ' s professional guidelines as expressed in the Comenius' Oath. This research contributes to the AI research by shedding light on how people within the HEI evaluate the impacts of AI into their future operating environment, pointing out also the potential obstacles for AI adoption in this specific context.
Python has made significant progress and its adaptability allows it to be utilized in nearly all fields recognized by humanity. The utilization of this technology in the field of research and development has left a di...
Python has made significant progress and its adaptability allows it to be utilized in nearly all fields recognized by humanity. The utilization of this technology in the field of research and development has left a distinctive impact. Advancements in brain research have occurred at a rapid pace, resulting in a continuous stream of novel results being unveiled on a daily basis. Electroencephalogram (EEG) signals refer to the electrical impulses generated by the brain in reaction to specific visual stimuli, such as positive, negative, neutral, or soothing stimuli. The utilization of Python, along with its robust modules and libraries, facilitates the efficient identification and subsequent study of human emotions, hence enabling emotion recognition. In this discussion, we will explore Long Short Term memory a machinelearning (ML) models that demonstrate efficacy in extracting and conveying significant insights pertaining to the human brain and EEG signals.
Heart patients are increasing in number on a daily basis, and each year in every region of the globe, a significant number of people lose their lives unexpectedly as a result of sudden heart *** a result of this, it i...
Heart patients are increasing in number on a daily basis, and each year in every region of the globe, a significant number of people lose their lives unexpectedly as a result of sudden heart *** a result of this, it is essential to diagnose cardiovascular illness at an early stage in order to avert the death of It is necessary for the area of medicine to make use of some kind of technology-based application in order to identify cardiac patients with greater precision and in much shorter time. It is possible to effectively identify heart illness by making use of datamining methods, and there are a considerable number of people suffering from heart disease who are now being treated in hospitals all over the globe. datamining is the practice of sifting through enormous amounts of raw data in search of knowledge or information that can be put to good use. During the prediction analysis, many machinelearning algorithms are put to use in order to locate useful patterns and make accurate projections on forthcoming events or trends. Through the use of a reworked machinelearning algorithm, this research study will attempt to forecast the risk that individuals would suffer from coronary heart disease. Before being classified, the incoming data go through a number of processes, including preprocessing, grouping, and the selection of useful features. Four different algorithms, including random forest (RF), K Nearest Neighbor (KNN), Particle swarm optimization (PSO) and logistic regression are combined in order to find out what kind of cardiac disease the patient *** enhance performance and shorten the amount of time needed for training, this method removes features from the cardiac dataset that are not relevant to the problem at hand. The random forest approach is the one that finishes off this procedure. In order to group all of the data points that are considered to be outliers, KNN are optimized using PSO algorithm. In the end, a method known as logistic regres
With the growing number of social and online sales platforms, the comments and reviews posted on these platforms play a significant role in choosing products and services. Nowadays, consumers actively search for and r...
详细信息
ISBN:
(数字)9798331516321
ISBN:
(纸本)9798331516338
With the growing number of social and online sales platforms, the comments and reviews posted on these platforms play a significant role in choosing products and services. Nowadays, consumers actively search for and read reviews left by other customers on online shopping sites before purchasing a particular product or service. And product manufacturers receive valuable feedback on the quality of their goods and services. Therefore, it is very important to take your place in the market and maintain it in a period when technologies are developing and rapidly evolving every day. Organizations do not need to conduct typical surveys to collect public opinion on their products and services, as this information is now available in open sources. To extract valuable information from customer reviews, opinion and sentiment analysis methods based on natural language processing are used. Opinions are divided into two types: non-comparative (regular) and comparative. Using advanced Natural Language Processing methods and machine and deep learning models, it is possible to automate the collection and analysis of comparative opinions, allowing businesses and researchers to make decisions based on comparable data. This paper provides a review of comparative opinion mining and sentiment analysis in the Kazakh language. Due to the limited availability of articles specifically addressing comparative opinion mining in Kazakh, we broadened our review to include related research in sentiment analysis overall. Additionally, the article discusses the structure of comparative sentences in both English and Kazakh, providing insights into linguistic frameworks.
Heatwaves are a serious challenge in India, claiming many lives yearly. The heatwaves occur when the temperature rises above the normal continuously for a few days. This paper proposes a Stacked LSTM model that predic...
详细信息
ISBN:
(数字)9798331527549
ISBN:
(纸本)9798331527556
Heatwaves are a serious challenge in India, claiming many lives yearly. The heatwaves occur when the temperature rises above the normal continuously for a few days. This paper proposes a Stacked LSTM model that predicts the maximum temperature. The study region is Akola in Maharashtra, India which is prone to heatwaves. It is observed that the Proposed Method gives less Mean Absolute Error as compared to other techniques like Multiple Linear Regression, Random Forest, and Support Vector Regression. After data collection, appropriate data pre-processing technique is applied. Model is built considering the pre-processed data which does not have any anomalies. The reason for improved result of Proposed model is its ability to handle the complex pattern present in the data and learn from it to give better predictions. Maximum temperature is selected as the target variable for prediction and the other features are humidity, dew point etc. are considered here.
This study focuses on the detection of pre-crime events in videos, specifically shoplifting. In video understanding, visual features, pose, and emotion information are inherently frame-centric, providing insights in u...
详细信息
ISBN:
(数字)9798350372977
ISBN:
(纸本)9798350372984
This study focuses on the detection of pre-crime events in videos, specifically shoplifting. In video understanding, visual features, pose, and emotion information are inherently frame-centric, providing insights in understanding within isolated frames. Our approach introduces a novel method by extracting activity information between frames and incorporating visual features augmented with human pose and emotion information to enhance the understanding of pre-shoplifting behaviors. We used a set of CCTV videos in which at the end of some clips the customers shoplift and in others they do not. We used these videos as training data with a transformer machinelearning architecture. We augmented low level video analysis with pose, and emotion information on frame level and extracted activity information between frames. We conducted a systematic series of experiments to assess the performance of our model under two distinct configurations. The first configuration involved frame-centric visual features augmented with pose and emotion information, while the second configuration focused on activity information between frames. The results demonstrate improvements, underscoring the pivotal role of extracting temporal dynamics between frames in capturing nuanced activities leading up to the crime. The comprehension and capture of these activity patterns between frames proved to be crucial for a more thorough examination of pre-shoplifting events. The essence of our study's contribution is hinged upon the utilization of extracting temporal dynamics between frames, enabling the capture of crucial activities of individuals prior to engaging in shoplifting.
Financial statement fraud detection is the process of identifying falsified financial statements. Traditional auditing methods are time-consuming, expensive, and subject to error. Therefore, adopting an efficient and ...
Financial statement fraud detection is the process of identifying falsified financial statements. Traditional auditing methods are time-consuming, expensive, and subject to error. Therefore, adopting an efficient and robust machinelearning mechanism is important. Unfortunately, the current data sources suffer from a severe class imbalance. The lack of sufficient fraudulent financial statement records inspires the use of various resampling techniques. This paper a) examines the efficiency of different resampling strategies to detect fraudulent financial statements while employing multi-layer feedforward neural networks, support vector machines, and naïve Bayes machinelearning models, and b) investigates the superiority of using Raw Accounting Variables (RAVs) over financial ratios for financial statement fraud detection. A benchmark dataset of numerical financial variables (RAVs and financial ratios) is used as features for model evaluation. The fraud labels correspond to the Accounting and Auditing Enforcement Releases by the U.S. Securities and Exchange Commission (SEC). We analyze the performance of the models on 28 RAVs and 14 financial ratios suggested by accounting experts. Using the area under the receiver operating characteristic curve (AUC) as the performance metric, the synthetic minority oversampling technique (SMOTE), along with a three-layer feedforward neural network (AUC: 0.863), greatly outperformed the RUSBoost (AUC: 0.717) model.
Credit card fraud (CCF) is a persistent issue in the financial sector with serious consequences. datamining has proven to be extremely useful in detecting fraud in online transactions. However, detecting CCF through ...
Credit card fraud (CCF) is a persistent issue in the financial sector with serious consequences. datamining has proven to be extremely useful in detecting fraud in online transactions. However, detecting CCF through datamining is quite a difficult task because of two causes: constant changes in the profiles of normal and fraudulent behaviour, and the highly skewed nature of the data sets. The outcome of fraud detection in credit card transactions depends on the sampling approach, detection techniques, and variable selection. This work studies the performance of K-Nearest Neighbor, Naive Bayes, Logistic Regression and Random Forest algorithms on a highly skewed dataset. The dataset contains 2,84,807 transactions and has been collected from European cardholder transactions. A hybrid of under-sampling and oversampling techniques has been used on the skewed data. The four techniques were utilized on both data namely preprocessed and raw, and the results are evaluated using specificity, accuracy, sensitivity, and F1-score. The outcomes show that the optimal accuracy for Naive Bayes, Logistic Regression, K-Nearest Neighbor and Random Forest classifiers are 98.72%, 52.34%, 96.89%, 91.67%, respectively. The comparative results indicate that K-Nearest Neighbor performs better than Logistic Regression, Random Forest and Naive Bayes techniques.
暂无评论