In this paper, we explore the multi-classification problem of acupuncture acupoints bas</span><span><span style="font-family:Verdana;">ed on </span><span style="font-family:Ve...In this paper, we explore the multi-classification problem of acupuncture acupoints bas</span><span><span style="font-family:Verdana;">ed on </span><span style="font-family:Verdana;">Bert</span><span style="font-family:Verdana;"> model, </span><i><span style="font-family:Verdana;">i.e.</span></i><span style="font-family:Verdana;">, we try to recommend the best main acupuncture point for treating the disease by classifying and predicting the main acupuncture point for the disease, and further explore its acupuncture point grouping to provide the medical practitioner with the optimal solution for treating the disease and improv</span></span></span><span style="font-family:Verdana;">ing</span><span style="font-family:""><span style="font-family:Verdana;"> the clinical decision-making ability. The Bert-Chinese-Acupoint model was constructed by retraining </span><span style="font-family:Verdana;">on the basis of</span><span style="font-family:Verdana;"> the Bert model, and the semantic features in terms of acupuncture points were added to the acupunctu</span></span><span style="font-family:""><span style="font-family:Verdana;">re point corpus in the fine-tuning process to increase the semantic features in terms of acupuncture </span><span style="font-family:Verdana;">points,</span><span style="font-family:Verdana;"> and compared with the machine learning method. The results show that the Bert-Chinese Acupoint model proposed in this paper has a 3% improvement in accuracy compared to the </span><span style="font-family:Verdana;">best performing</span><span style="font-family:Verdana;"> model in the machine learning approach.展开更多
After a century of relative stability in the electricity sector,the widespread adoption of distributed energy resources,along with recent advancements in computing and communication technologies,has fundamentally alte...After a century of relative stability in the electricity sector,the widespread adoption of distributed energy resources,along with recent advancements in computing and communication technologies,has fundamentally altered how energy is consumed,traded,and utilized.This change signifies a crucial shift as the power system evolves from its traditional hierarchical organization to a more decentralized approach.At the heart of this transformation are innovative energy distribution models,like peer-to-peer(P2P)sharing,which enable communities to collaboratively manage their energy resources.The effectiveness of P2P sharing not only improves the economic prospects for prosumers,who generate and consume energy,but also enhances energy resilience and sustainability.This allows communities to better leverage local resources while fostering a sense of collective responsibility and collaboration in energy management.However,there is still no extensive implementation of such sharing models in today’s electricitymarkets.Research on distributed energy P2P trading is still in the exploratory stage,and it is particularly important to comprehensively understand and analyze the existing distributed energy P2P trading market.This paper contributes with an overview of the P2P markets that starts with the network framework,market structure,technical approach for trading mechanism,and blockchain technology,moving to the outlook in this field.展开更多
There has been evidence of crime in the US since colonization. In this article, we analyze the crime statistics of San Francisco and its resolution of crime recorded from January to September of the year 2018. We defi...There has been evidence of crime in the US since colonization. In this article, we analyze the crime statistics of San Francisco and its resolution of crime recorded from January to September of the year 2018. We define resolution of crime as a target variable and study its relationship with other variables. We make several classification models to predict resolution of crime using several data mining techniques and suggest the best model for predicting resolution.展开更多
Cardiovascular disease (CVD) is a leading cause of death across the globe. Approximately 17.9 million of people die globally each year due to CVD, </span><span style="font-family:Verdana;">which ...Cardiovascular disease (CVD) is a leading cause of death across the globe. Approximately 17.9 million of people die globally each year due to CVD, </span><span style="font-family:Verdana;">which comprises 31% of all death. Coronary Artery Disease (CAD) is a common</span><span style="font-family:Verdana;"> type of CVD and is considered fatal.</span></span><span style="font-family:""> </span><span style="font-family:Verdana;">Predictive models that use machine learning algorithms may assist health workers in timely detection of CAD which ultimately reduce</span><span style="font-family:Verdana;">s</span><span style="font-family:Verdana;"> the mortality.</span><span style="font-family:""> </span><span style="font-family:""><span style="font-family:Verdana;">The main purpose of this study is to build a predictive model that provides doctors and health care providers with personalized information to implement better and more personalized treat</span><span style="font-family:Verdana;">ments for their patients. In</span></span><span style="font-family:""> </span><span style="font-family:Verdana;">this study, we use the publicly available Z-Alizadeh</span><span style="font-family:Verdana;"> Sani dataset which contains random samples of 216 cases with CAD and 87 normal controls with 56 different features. The binary variable “Cath” which represents case-control status, is used the target variable. We study its relationship with other predictors and develop classification models using the five different supervised classification machine learning algorithms: Logistic Regression (LR), Classification Tree</span><span style="font-family:""> </span><span style="font-family:Verdana;">with</span><span style="font-family:""> </span><span style="font-family:""><span style="font-family:Verdana;">Bagging (Bagging CART), </span><span style="font-family:Verdana;">Random </span><span style="font-family:Verdana;">Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbors (KNN).</span><span style="font-family:Verdana;"> These five classification models are used to investigate the detection of CAD. Finally, the performance of the machine learning algorithms is compared,</span></span><span style="font-family:""> </span><span style="font-family:Verdana;">and the best model is selected. Our results indicate that the SVM model is able to predict the presence of CAD more effectively and accurately than other models with an accuracy of 0.8947, sensitivity of 0.9434, specificity of 0.7826, and AUC of 0.8868.展开更多
文摘In this paper, we explore the multi-classification problem of acupuncture acupoints bas</span><span><span style="font-family:Verdana;">ed on </span><span style="font-family:Verdana;">Bert</span><span style="font-family:Verdana;"> model, </span><i><span style="font-family:Verdana;">i.e.</span></i><span style="font-family:Verdana;">, we try to recommend the best main acupuncture point for treating the disease by classifying and predicting the main acupuncture point for the disease, and further explore its acupuncture point grouping to provide the medical practitioner with the optimal solution for treating the disease and improv</span></span></span><span style="font-family:Verdana;">ing</span><span style="font-family:""><span style="font-family:Verdana;"> the clinical decision-making ability. The Bert-Chinese-Acupoint model was constructed by retraining </span><span style="font-family:Verdana;">on the basis of</span><span style="font-family:Verdana;"> the Bert model, and the semantic features in terms of acupuncture points were added to the acupunctu</span></span><span style="font-family:""><span style="font-family:Verdana;">re point corpus in the fine-tuning process to increase the semantic features in terms of acupuncture </span><span style="font-family:Verdana;">points,</span><span style="font-family:Verdana;"> and compared with the machine learning method. The results show that the Bert-Chinese Acupoint model proposed in this paper has a 3% improvement in accuracy compared to the </span><span style="font-family:Verdana;">best performing</span><span style="font-family:Verdana;"> model in the machine learning approach.
基金funded by the National Natural Science Foundation of China(52167013)the Key Program of Natural Science Foundation of Gansu Province(24JRRA225)Natural Science Foundation of Gansu Province(23JRRA891).
文摘After a century of relative stability in the electricity sector,the widespread adoption of distributed energy resources,along with recent advancements in computing and communication technologies,has fundamentally altered how energy is consumed,traded,and utilized.This change signifies a crucial shift as the power system evolves from its traditional hierarchical organization to a more decentralized approach.At the heart of this transformation are innovative energy distribution models,like peer-to-peer(P2P)sharing,which enable communities to collaboratively manage their energy resources.The effectiveness of P2P sharing not only improves the economic prospects for prosumers,who generate and consume energy,but also enhances energy resilience and sustainability.This allows communities to better leverage local resources while fostering a sense of collective responsibility and collaboration in energy management.However,there is still no extensive implementation of such sharing models in today’s electricitymarkets.Research on distributed energy P2P trading is still in the exploratory stage,and it is particularly important to comprehensively understand and analyze the existing distributed energy P2P trading market.This paper contributes with an overview of the P2P markets that starts with the network framework,market structure,technical approach for trading mechanism,and blockchain technology,moving to the outlook in this field.
文摘There has been evidence of crime in the US since colonization. In this article, we analyze the crime statistics of San Francisco and its resolution of crime recorded from January to September of the year 2018. We define resolution of crime as a target variable and study its relationship with other variables. We make several classification models to predict resolution of crime using several data mining techniques and suggest the best model for predicting resolution.
文摘Cardiovascular disease (CVD) is a leading cause of death across the globe. Approximately 17.9 million of people die globally each year due to CVD, </span><span style="font-family:Verdana;">which comprises 31% of all death. Coronary Artery Disease (CAD) is a common</span><span style="font-family:Verdana;"> type of CVD and is considered fatal.</span></span><span style="font-family:""> </span><span style="font-family:Verdana;">Predictive models that use machine learning algorithms may assist health workers in timely detection of CAD which ultimately reduce</span><span style="font-family:Verdana;">s</span><span style="font-family:Verdana;"> the mortality.</span><span style="font-family:""> </span><span style="font-family:""><span style="font-family:Verdana;">The main purpose of this study is to build a predictive model that provides doctors and health care providers with personalized information to implement better and more personalized treat</span><span style="font-family:Verdana;">ments for their patients. In</span></span><span style="font-family:""> </span><span style="font-family:Verdana;">this study, we use the publicly available Z-Alizadeh</span><span style="font-family:Verdana;"> Sani dataset which contains random samples of 216 cases with CAD and 87 normal controls with 56 different features. The binary variable “Cath” which represents case-control status, is used the target variable. We study its relationship with other predictors and develop classification models using the five different supervised classification machine learning algorithms: Logistic Regression (LR), Classification Tree</span><span style="font-family:""> </span><span style="font-family:Verdana;">with</span><span style="font-family:""> </span><span style="font-family:""><span style="font-family:Verdana;">Bagging (Bagging CART), </span><span style="font-family:Verdana;">Random </span><span style="font-family:Verdana;">Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbors (KNN).</span><span style="font-family:Verdana;"> These five classification models are used to investigate the detection of CAD. Finally, the performance of the machine learning algorithms is compared,</span></span><span style="font-family:""> </span><span style="font-family:Verdana;">and the best model is selected. Our results indicate that the SVM model is able to predict the presence of CAD more effectively and accurately than other models with an accuracy of 0.8947, sensitivity of 0.9434, specificity of 0.7826, and AUC of 0.8868.
