المؤلفون: Lian, Shiguo, Zhao, Kaikai, Liu, Xinhui, Lei, Xuejiao, Yang, Bikun, Zhang, Wenjing, Wang, Kai, Liu, Zhaoxiang

مصطلحات موضوعية: Computer Science - Computation and Language, Computer Science - Artificial Intelligence

الوصف: General large language models enhanced with supervised fine-tuning and reinforcement learning from human feedback are increasingly popular in academia and industry as they generalize foundation models to various practical tasks in a prompt manner. To assist users in selecting the best model in practical application scenarios, i.e., choosing the model that meets the application requirements while minimizing cost, we introduce A-Eval, an application-driven LLMs evaluation benchmark for general large language models. First, we categorize evaluation tasks into five main categories and 27 sub-categories from a practical application perspective. Next, we construct a dataset comprising 678 question-and-answer pairs through a process of collecting, annotating, and reviewing. Then, we design an objective and effective evaluation method and evaluate a series of LLMs of different scales on A-Eval. Finally, we reveal interesting laws regarding model scale and task difficulty level and propose a feasible method for selecting the best model. Through A-Eval, we provide clear empirical and engineer guidance for selecting the best model, reducing barriers to selecting and using LLMs and promoting their application and development. Our benchmark is publicly available at https://github.com/UnicomAI/DataSet/tree/main/TestData/GeneralAbilityTest.

الوصول الحر: http://arxiv.org/abs/2406.10307Test

المؤلفون: Liu, Xinhui, Chen, Zhenghao, Zhou, Luping, Xu, Dong, Xi, Wei, Bai, Gairui, Zhao, Yihan, Zhao, Jizhong

المصدر: AAAI2024

مصطلحات موضوعية: Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition

الوصف: Conventional Federated Domain Adaptation (FDA) approaches usually demand an abundance of assumptions, which makes them significantly less feasible for real-world situations and introduces security hazards. This paper relaxes the assumptions from previous FDAs and studies a more practical scenario named Universal Federated Domain Adaptation (UFDA). It only requires the black-box model and the label set information of each source domain, while the label sets of different source domains could be inconsistent, and the target-domain label set is totally blind. Towards a more effective solution for our newly proposed UFDA scenario, we propose a corresponding methodology called Hot-Learning with Contrastive Label Disambiguation (HCLD). It particularly tackles UFDA's domain shifts and category gaps problems by using one-hot outputs from the black-box models of various source domains. Moreover, to better distinguish the shared and unknown classes, we further present a cluster-level strategy named Mutual-Voting Decision (MVD) to extract robust consensus knowledge across peer classes from both source and target domains. Extensive experiments on three benchmark datasets demonstrate that our method achieves comparable performance for our UFDA scenario with much fewer assumptions, compared to previous methodologies with comprehensive additional assumptions.
Comment: Accepted by AAAI2024

الوصول الحر: http://arxiv.org/abs/2311.15570Test

المؤلفون: Zhao, Puyang, Tian, Wei, Xiao, Lefu, Liu, Xinhui, Wu, Jingjin

مصطلحات موضوعية: Computer Science - Cryptography and Security, Computer Science - Computers and Society, Computer Science - Machine Learning

الوصف: Bitcoin is the most common cryptocurrency involved in cyber scams. Cybercriminals often utilize pseudonymity and privacy protection mechanism associated with Bitcoin transactions to make their scams virtually untraceable. The Ponzi scheme has attracted particularly significant attention among Bitcoin fraudulent activities. This paper considers a multi-class classification problem to determine whether a transaction is involved in Ponzi schemes or other cyber scams, or is a non-scam transaction. We design a specifically designed crawler to collect data and propose a novel Attention-based Long Short-Term Memory (A-LSTM) method for the classification problem. The experimental results show that the proposed model has better efficiency and accuracy than existing approaches, including Random Forest, Extra Trees, Gradient Boosting, and classical LSTM. With correctly identified scam features, our proposed A-LSTM achieves an F1-score over 82% for the original data and outperforms the existing approaches.

الوصول الحر: http://arxiv.org/abs/2210.14408Test

المؤلفون: ZHOU Xiaoqi, LIU Xinhui, ZHANG Wei, LI Changfeng, YAN Yaqiong

المصدر: Zhongguo quanke yixue, Vol 26, Iss 29, Pp 3645-3649 (2023)

مصطلحات موضوعية: diabetes mellitus, type 2, metabolic syndrome, alanine transaminase, aspartate aminotransferase, aged, correlation study, Medicine

الوصف: Background Serum alanine transaminase (ALT) and aspartate aminotransferase (AST) are common liver enzymes, but there are few studies on the correlation of these enzymes with the prevalence of type 2 diabetes mellitus (T2DM) and metabolic syndrome (MS) . Objective To explore the association of serum ALT and AST/ALT ratio with T2DM and MS in older physical examinees in the community. Methods From January to December, 2018, 30 060 elderly people (≥65 years) who underwent free physical examination in 19 community health centers of 6 districts in Wuhan were chosen as the subjects. Their demographic data, life style, previous medical history, and results of physical examination and biochemical test were collected. Subjects were divided into quartile groups of ALT〔Q1 (ALT≤11 U/L, n=8 116), Q2 (11 U/L19 U/L, n=7 498) 〕, or quartile groups of AST/ALT ratio〔q1 (AST/ALT≤1.06, n=7 386), q2 (1.061.62, n=7 447) 〕. Cochran χ2 test was used to compare the prevalence trend of T2DM, MS, abdominal obesity, hypertension, elevated triglycerides (TG) and lowered high-density lipoprotein cholesterol (HDL-C) with the change of ALT level and AST/ALT ratio. Multivariate Logistic regression analysis was used to explore the correlation of prevalence of T2DM and MS with ALT level and AST/ALT ratio. Results The age of 30 060 subjects (13 381 men and 16 679 women) was 65-99 years old, and the average age was (71.7±5.6) years old. The prevalence of T2DM and MS was 18.76% (5 488/30 060) and 29.73% (8 938/30 060), respectively. The regular exercise rate, BMI, waist circumference, systolic pressure, diastolic pressure, ALT, TG and fasting plasma glucose (FPG) were significantly higher and smoking rate, drinking rate, AST, AST/ALT ratio, total cholesterol (TC), HDL-C, low-density lipoprotein cholesterol (LDL-C) were significantly lower in T2DM subjects compared to non-T2DM subjects (P

وصف الملف: electronic resource

العلاقة: https://www.chinagp.net/fileup/1007-9572/PDF/zx20230028.pdfTest; https://doaj.org/toc/1007-9572Test

الوصول الحر: https://doaj.org/article/eaf2ceeb39fd4d4294a202869ec23efdTest

المؤلفون: Lin, Tianrui, Wei, Tianjian, Hu, Qingqing, Fu, Chunfang, Huang, Nuo, Liu, Xinhui, Tang, Li, Su, Liang, Luo, Jianghua, Gong, Chen

المساهمون: National Natural Science Foundation of China

المصدر: IEEE Photonics Journal ; volume 16, issue 4, page 1-10 ; ISSN 1943-0655 1943-0647

الإتاحة: https://doi.org/10.1109/jphot.2024.3407876Test
http://xplorestaging.ieee.org/ielx8/4563994/10552336/10543051.pdf?arnumber=10543051Test

المؤلفون: Krys, Kuba, Chun Yeung, June, Haas, Brian W., van Osch, Yvette, Kosiarczyk, Aleksandra, Kocimska-Zych, Agata, Torres, Claudio, Selim, Heyla A., Zelenski, John M., Bond, Michael Harris, Park, Joonha, Lun, Vivian Miu Chi, Maricchiolo, Fridanna, Vauclair, Christin Melanie, Poláčková Šolcová, Iva, Sirlopú, David, Xing, Cai, Vignoles, Vivian L., van Tilburg, Wijnand A.P., Teyssier, Julien, Sun, Chien Ru, Serdarevich, Ursula, Schwarz, Beate, Sargautyte, Ruta, Røysamb, Espen, Romashov, Vladyslav, Rizwan, Muhammad, Pavlović, Zoran, Pavlopoulos, Vassilis, Okvitawanli, Ayu, Nadi, Azar, Nader, Martin, Mustaffa, Nur Fariza, Murdock, Elke, Mosca, Oriana, Mohorić, Tamara, Barrientos Marroquin, Pablo Eduardo, Malyonova, Arina, Liu, Xinhui, Lee, J. Hannah, Kwiatkowska, Anna, Kronberger, Nicole, Klůzová Kráčmarová, Lucie, Kascakova, Natalia, Işık, İdil, Igou, Eric R., Igbokwe, David O., Hanke-Boer, Diana, Gavreliuc, Alin, Garðarsdóttir, Ragna B., Fülöp, Márta, Gamsakhurdia, Vladimer, Esteves, Carla Sofia, Domínguez-Espinosa, Alejandra, Denoux, Patrick, Charkviani, Salome, Baltin, Arno, Arevalo, Douglas, Appoh, Lily, Akotia, Charity, Adamovic, Mladen, Uchida, Yukiko

مصطلحات موضوعية: Culture, Family, Happiness, Interdependent happiness, Life satisfaction, Relational mobility, Well-being

الوصف: People care about their own well-being and about the well-being of their families. It is currently, however, unknown how much people tend to value their own versus their family’s well-being. A recent study documented that people value family happiness over personal happiness across four cultures. In this study, we sought to replicate this finding across a larger sample size (N = 12,819) and a greater number of countries (N = 49). We found that the strength of the idealization of family over personal happiness preference was small (average Cohen’s ds =.20, range −.02 to.48), but present in 98% of the studied countries, with statistical significance in 73% to 75%, and variance across countries <2%. We also found that the size of this effect did vary somewhat across cultural contexts. In Latin American cultures highest on relational mobility, the idealization of family over personal happiness was very small (average Cohen’s ds for Latin America =.15 and.18), while in Confucian Asia cultures lowest on relational mobility, this effect was closer to medium (ds >.40 and.30). Importantly, we did not find strong support for traditional theories in cross-cultural psychology that associate collectivism with greater prioritization of the family versus the individual; country-level individualism–collectivism was not associated with variation in the idealization of family versus individual happiness. Our findings indicate that no matter how much various populists abuse the argument of “protecting family life” to disrupt emancipation, family happiness seems to be a pan-culturally phenomenon. Family well-being is a key ingredient of social fabric across the world, and should be acknowledged by psychology and well-being researchers and by progressive movements too. ; info:eu-repo/semantics/acceptedVersion

العلاقة: http://hdl.handle.net/10400.14/44844Test; 85149537802; 000956272100001

الإتاحة: https://doi.org/10.1177/00220221221134711Test
http://hdl.handle.net/10400.14/44844Test

المؤلفون: Zhao, Puyang, Liu, Xinhui, Yue, Zhiyi, Zhao, Qianyu, Liu, Xinzhi, Deng, Yuhui, Wu, Jingjin

مصطلحات موضوعية: R Medicine, QA75 Electronic computers. Computer science

الوصف: In the rapidly evolving field of medical diagnostics, the challenge of imbalanced datasets, particularly in diabetes classification, calls for innovative solutions. The study introduces DiGAN, a groundbreaking approach that leverages the power of Generative Adversarial Networks (GAN) to revolutionize diabetes data analysis. Marking a significant departure from traditional methods, DiGAN applies GANs, typically seen in image processing, to the realm of diabetes data. This novel application is complemented by integrating the unsupervised Laplacian Score for sophisticated feature selection. The pioneering approach not only surpasses the limitations of existing techniques but also sets a new benchmark in classification accuracy with a 90% weighted F1-score, achieving a remarkable improvement of over 20% compared to conventional methods. Additionally, DiGAN demonstrates superior performance over popular SMOTE-based methods in handling extremely imbalanced datasets. This research, focusing on the integrated use of Laplacian Score, GAN, and Random Forest, stands at the forefront of diabetic classification, offering a uniquely effective and innovative solution to the long-standing data imbalance issue in medical diagnostics.

وصف الملف: text

العلاقة: http://eprints.lse.ac.uk/122841/1/Liu_digan_breakthrough_published.pdfTest; Zhao, Puyang, Liu, Xinhui, Yue, Zhiyi, Zhao, Qianyu, Liu, Xinzhi, Deng, Yuhui and Wu, Jingjin (2024) DiGAN breakthrough: advancing diabetic data analysis with innovative GAN-based imbalance correction techniques. Computer Methods and Programs in Biomedicine Update, 5. ISSN 2666-9900

الإتاحة: http://eprints.lse.ac.uk/122841Test/
http://eprints.lse.ac.uk/122841/1/Liu_digan_breakthrough_published.pdfTest
https://www.sciencedirect.com/journal/computer-methods-and-programs-in-biomedicine-updateTest

المؤلفون: Zhang, Yueqi, Luo, Yue, Liu, Xinhui, Kiupel, Matti, Li, Aimin, Wang, Hongbing, Mi, Qing-Sheng, Xiao, Hua

المصدر: Cellular and Molecular Gastroenterology and Hepatology ; volume 17, issue 1, page 1-27 ; ISSN 2352-345X

مصطلحات موضوعية: Gastroenterology, Hepatology

الإتاحة: https://doi.org/10.1016/j.jcmgh.2023.09.007Test
https://api.elsevier.com/content/article/PII:S2352345X23001698?httpAccept=text/xmlTest
https://api.elsevier.com/content/article/PII:S2352345X23001698?httpAccept=text/plainTest

المؤلفون: Krys, Kuba, Haas, Brian W., Igou, Eric Raymond, Kosiarczyk, Aleksandra, Kocimska-Bortnowska, Agata, Kwiatkowska, Anna, Lun, Vivian Miu-Chi, Maricchiolo, Fridanna, Park, Joonha, Šolcová, Iva Poláčková, Sirlopú, David, Uchida, Yukiko, Vauclair, Christin-Melanie, Vignoles, Vivian L., Zelenski, John M., Adamovic, Mladen, Akotia, Charity S., Albert, Isabelle, Appoh, Lily, Mira, D. M. Arévalo, Baltin, Arno, Denoux, Patrick, Domínguez-Espinosa, Alejandra, Esteves, Carla Sofia, Gamsakhurdia, Vladimer, Fülöp, Márta, Garðarsdóttir, Ragna B., Gavreliuc, Alin, Boer, Diana, Igbokwe, David O., Işık, İdil, Kascakova, Natalia, Klůzová Kráčmarová, Lucie, Kosakowska-Berezecka, Natasza, Kostoula, Olga, Kronberger, Nicole, Lee, J. Hannah, Liu, Xinhui, Łużniak-Piecha, Magdalena, Malyonova, Arina, Barrientos, Pablo Eduardo, Mohorić, Tamara, Mosca, Oriana, Murdock, Elke, Mustaffa, Nur Fariza, Nader, Martin, Nadi, Azar, Okvitawanli, Ayu, van Osch, Yvette, Pavlopoulos, Vassilis, Pavlović, Zoran, Rizwan, Muhammad, Romashov, Vladyslav, Røysamb, Espen, Sargautytė, Rūta, Schwarz, Beate, Selim, Heyla A., Serdarevich, Ursula, Stogianni, Maria, Sun, Chien-Ru, Teyssier, Julien, van Tilburg, Wijnand A. P., Torres, Claudio, Xing, Cai, Bond, Michael Harris

المصدر: Journal of happiness studies., Dordrecht : Springer Science and Business Media B.V., 2023, vol. 24, iss. 2, p. 607-627. ; ISSN 1389-4978 ; eISSN 1573-7780

مصطلحات موضوعية: cultural sensitivity, culture, happiness, interdependent happiness, life satisfaction, self-construals, selfhoods, well-being

الوصف: How can one conclude that well-being is higher in country A than country B, when well-being is being measured according to the way people in country A think about well-being? We address this issue by proposing a new culturally sensitive method to comparing societal levels of well-being. We support our reasoning with data on life satisfaction and interdependent happiness focusing on individual and family, collected mostly from students, across forty-nine countries. We demonstrate that the relative idealization of the two types of well-being varies across cultural contexts and are associated with culturally different models of selfhood. Furthermore, we show that rankings of societal well-being based on life satisfaction tend to underestimate the contribution from interdependent happiness. We introduce a new culturally sensitive method for calculating societal well-being, and examine its construct validity by testing for associations with the experience of emotions and with individualism-collectivism. This new culturally sensitive approach represents a slight, yet important improvement in measuring well-being.

وصف الملف: application/pdf

العلاقة: https://epublications.vu.lt/object/elaba:159248208/159248208.pdfTest; https://repository.vu.lt/VU:ELABAPDB159248208&prefLang=en_USTest

الإتاحة: https://doi.org/10.1007/s10902-022-00588-1Test
https://repository.vu.lt/VU:ELABAPDB159248208&prefLang=en_USTest

المؤلفون: Sánchez-Rodríguez, Ángel, Vignoles, Vivian L, Bond, Michael Harris, Adamovic, Mladen, Akotia, Charity S., Albert, Isabelle, Appoh, Lily, Baltin, Arno, Barrientos, Pablo Eduardo, Denoux, Patrick, Domínguez-Espinosa, Alejandra, Esteves, Carla Sofia, Fülöp, Márta, Gamsakhurdia, Vladimer, Gardarsdottir, Ragna B, Gavreliuc, Alin, Hanke-Boer, Diana, Haas, Brian W., Igbokwe, David O., Isik, Idil, Kaščáková, Natália, Kračmárová, Lucie Klůzová, Kocimska-Zych, Agata, Kosiarczyk, Aleksandra, Kostoula, Olga, Kronberger, Nicole, Krys, Kuba, Kwiatkowska, Anna, Lee, J. Hannah, Liu, Xinhui, Łużniak-Piecha, Magdalena, Маленова, Арина, Fridanna, MARICCHIOLO, Mira, Douglas Marlon Arévalo, Mohoric, Tamara, Mosca, Oriana, Murdock, Elke, Mustaffa, Nur Fariza, Lun, Vivian, Nader, Martin, Nadi, Azar, Okvitawanli, Ayu, Osch, Yvette van, Park, Joonha, Pavlopoulos, Vassilis, Pavlović, Zoran, Šolcová, Iva Poláčková, Igou, Eric R., Rizwan, Muhammad, Romashov, Vladyslav, Røysamb, Espen, Sargautyte, Ruta, Schwarz, Beate, Selim, Heyla A., Serdarevich, Ursula, Sirlopu, David, Stogianni, Maria, Stoyanova, Stanislava, Sun, Chien-Ru, Teyssier, Julien, van Tilburg, Wijnand A. P., Torres, Claudio, Uchida, Yukiko, Vauclair, Christin-Melanie, Xing, Cai, Zelenski, John M.

المصدر: Sánchez-Rodríguez , Á , Vignoles , V L , Bond , M H , Adamovic , M , Akotia , C S , Albert , I , Appoh , L , Baltin , A , Barrientos , P E , Denoux , P , Domínguez-Espinosa , A , Esteves , C S , Fülöp , M , Gamsakhurdia , V , Gardarsdottir , R B , Gavreliuc , A , Hanke-Boer , D , Haas , B W , Igbokwe , D O , Isik , I , Kaščáková , N , Kračmárová , L K , Kocimska-Zych , ....

مصطلحات موضوعية: Economic inequality, Multicomponentself-construal, National wealth, Personal life satisfaction, Religious heritage

الوصف: We explore to what extent previously observed pan-cultural association between dimensions of self-construal and personal life satisfaction (PLS) may be moderated by three national-contextual variables: national wealth, economic inequality, and religious heritage. The results showed that Self-reliance (vs. dependence on others) predicted PLS positively in poorer countries but negatively in richer countries. Connectedness to others (vs. self-containment) predicted PLS more strongly in Protestant-heritage countries. Self-expression (vs. harmony) predicted PLS more weakly (and non-significantly) in Muslim-heritage countries. In contrast, previously reported associations of self-direction (vs. reception-to-influence), consistency (vs. variability), and decontextualized (vs. contextualized) self-understanding with personal life satisfaction were not significantly moderated by these aspects of societal context. These results show the importance of considering the impact of national religious and economic context.

العلاقة: https://research.tilburguniversity.edu/en/publications/045e461d-da27-4a87-b893-4cd609730155Test

الإتاحة: https://doi.org/10.1080/15298868.2023.2202413Test
https://research.tilburguniversity.edu/en/publications/045e461d-da27-4a87-b893-4cd609730155Test
http://www.scopus.com/inward/record.url?scp=85158162730&partnerID=8YFLogxKTest