Latent factor(LF)models are highly effective in extracting useful knowledge from High-Dimensional and Sparse(HiDS)matrices which are commonly seen in various industrial applications.An LF model usually adopts iterativ...Latent factor(LF)models are highly effective in extracting useful knowledge from High-Dimensional and Sparse(HiDS)matrices which are commonly seen in various industrial applications.An LF model usually adopts iterative optimizers,which may consume many iterations to achieve a local optima,resulting in considerable time cost.Hence,determining how to accelerate the training process for LF models has become a significant issue.To address this,this work proposes a randomized latent factor(RLF)model.It incorporates the principle of randomized learning techniques from neural networks into the LF analysis of HiDS matrices,thereby greatly alleviating computational burden.It also extends a standard learning process for randomized neural networks in context of LF analysis to make the resulting model represent an HiDS matrix correctly.Experimental results on three HiDS matrices from industrial applications demonstrate that compared with state-of-the-art LF models,RLF is able to achieve significantly higher computational efficiency and comparable prediction accuracy for missing data.I provides an important alternative approach to LF analysis of HiDS matrices,which is especially desired for industrial applications demanding highly efficient models.展开更多
OBJECTIVE: To explore the features of Traditional Chinese Medicine(TCM) syndromes in male infertility using computer-based analyses.METHODS: Latent class analysis was used to analyze the TCM syndrome data from 813 pat...OBJECTIVE: To explore the features of Traditional Chinese Medicine(TCM) syndromes in male infertility using computer-based analyses.METHODS: Latent class analysis was used to analyze the TCM syndrome data from 813 patients with male infertility and establish a latent tree model.RESULTS: A latent tree model with a Bayesian information criterion score of-11 263 was created.This model revealed that the characteristics of basic TCM syndromes in patients with male infertility were kidney Yang deficiency, kidney Qi deficiency,spleen Yang deficiency, liver Qi stagnation, Qi stagnation and blood stasis, and dump-heat; moreover,most patients with male infertility had complex syndromes(spleen-kidney Yang deficiency and liver Qi stagnation) rather than simple single syndromes.CONCLUSION: The hidden tree model analysis revealed the objective and quantitative complex relationships between the TCM symptoms of male infertility, and obtained the quantification and objective evidence of TCM syndromes in male infertility.展开更多
Mining stimulates environmental and economic impacts on the neighboring community right from the inception to the closure of its operations. The society in the neighborhood of mining gradually adopts a characteristic ...Mining stimulates environmental and economic impacts on the neighboring community right from the inception to the closure of its operations. The society in the neighborhood of mining gradually adopts a characteristic life-style that is highly influenced by the mining. In order to sustain the societal development beyond the mine closure, it is necessary to plan post mining activities in the area. Thus, it is essential to predict the impacts of mine closure well before the closure. Many societal and family attributes are affected by mine closure. Impact on these attributes is reflected on the overall quality of life of the neighboring community. There are no adequate indicators and/or methodology available to measure social impacts of mine closure on a neighboring community. This paper made an attempt to develop such methodology to predict the degree of adverse effects of mine closure on the quality of life of neighboring communities using the Structural Equation Modeling (SEM) and the Latent Variables Interaction Model (LVM).展开更多
Latent class analysis (LCA) is a widely used statistical technique for identifying subgroups in the population based upon multiple indicator variables. It has a number of advantages over other unsupervised grouping pr...Latent class analysis (LCA) is a widely used statistical technique for identifying subgroups in the population based upon multiple indicator variables. It has a number of advantages over other unsupervised grouping procedures such as cluster analysis, including stronger theoretical underpinnings, more clearly defined measures of model fit, and the ability to conduct confirmatory analyses. In addition, it is possible to ascertain whether an LCA solution is equally applicable to multiple known groups, using invariance assessment techniques. This study compared the effectiveness of multiple statistics for detecting group LCA invariance, including a chi-square difference test, a bootstrap likelihood ratio test, and several information indices. Results of the simulation study found that the bootstrap likelihood ratio test was the optimal invariance assessment statistic. In addition to the simulation, LCA group invariance assessment was demonstrated in an application with the Youth Risk Behavior Survey (YRBS). Implications of the simulation results for practice are discussed.展开更多
The climatological mean state, seasonal variation and long-term upward trend of 1979-2005 latent heat flux （LHF） in historical runs of 14 coupled general circulation models from CMIP5 （Coupled Model Intercomparison...The climatological mean state, seasonal variation and long-term upward trend of 1979-2005 latent heat flux （LHF） in historical runs of 14 coupled general circulation models from CMIP5 （Coupled Model Intercomparison Project Phase 5） are evaluated against OAFlux （Objectively Analyzed air-sea Fluxes） data. Inter-model diversity of these models in simulating the annual mean climatological LHF is discussed. Results show that the models can capture the climatological LHF fairly well, but the amplitudes are generally overestimated. Model-simulated seasonal variations of LHF match well with observations with overestimated amplitudes. The possible origins of these biases are wind speed biases in the CMIP5 models. Inter-model diversity analysis shows that the overall stronger or weaker LHF over the tropical and subtropical Pacific region, and the meridional variability of LHF, are the two most notable diversities of the CMIP5 models. Regression analysis indicates that the inter-model diversity may come from the diversity of simulated SST and near-surface atmospheric specific humidity. Comparing the observed long-term upward trend, the trends of LHF and wind speed are largely underestimated, while trends of SST and air specific humidity are grossly overestimated, which may be the origins of the model biases in reproducing the trend of LHF.展开更多
Background: Workplace violence (WV) towards psychiatric staff has commonly been associated with Posttraumatic Stress Disorder (PTSD). However, prospective studies have shown that not all psychiatric staff who experien...Background: Workplace violence (WV) towards psychiatric staff has commonly been associated with Posttraumatic Stress Disorder (PTSD). However, prospective studies have shown that not all psychiatric staff who experience workplace violence experience post-traumatic stress. Purpose: We want to examine the longitudinal trajectories of PTSD in this population to identify possible subgroups that might be more at risk. Furthermore, we need to investigate whether certain risk factors of PTSD might identify membership in the subgroups. Method: In a sample of psychiatric staff from 18 psychiatric wards in Denmark who had reported an incident of WV, we used Latent Growth Mixture Modelling (LGMM) and further logistic regression analysis to investigate this. Results: We found three separate PTSD trajectories: a recovering, a delayed-onset, and a moderate-stable trajectory. Higher social support and negative cognitive appraisals about oneself, the world and self-blame predicted membership in the delayed-onset trajectory, while higher social support and lower accept coping predicted membership in the delayed-onset trajectory. Conclusion: Although most psychiatric staff go through a natural recovery, it is important to be aware of and identify staff members who might be struggling long-term. More focus on the factors that might predict these groups should be an important task for psychiatric departments to prevent posttraumatic symptomatology from work.展开更多
In this project, we consider obtaining Fourier features via more efficient sampling schemes to approximate the kernel in LFMs. A latent force model (LFM) is a Gaussian process whose covariance functions follow an Expo...In this project, we consider obtaining Fourier features via more efficient sampling schemes to approximate the kernel in LFMs. A latent force model (LFM) is a Gaussian process whose covariance functions follow an Exponentiated Quadratic (EQ) form, and the solutions for the cross-covariance are expensive due to the computational complexity. To reduce the complexity of mathematical expressions, random Fourier features (RFF) are applied to approximate the EQ kernel. Usually, the random Fourier features are implemented with Monte Carlo sampling, but this project proposes replacing the Monte-Carlo method with the Quasi-Monte Carlo (QMC) method. The first-order and second-order models’ experiment results demonstrate the decrease in NLPD and NMSE, which revealed that the models with QMC approximation have better performance.展开更多
Latent variable models can effectively determine the condition of essential rotating machinery without needing labeled data.These models analyze vibration data via an unsupervised learning strategy.Temporal preservati...Latent variable models can effectively determine the condition of essential rotating machinery without needing labeled data.These models analyze vibration data via an unsupervised learning strategy.Temporal preservation is necessary to obtain an informative latent manifold for the fault diagnosis task.In a temporalpreserving context,two approaches exist to develop a condition-monitoring methodology:offline and online.For latent variable models,the available training modes are not different.While many traditional methods use offline training,online training can dynamically adjust the latent manifold,possibly leading to better fault signature extraction from the vibration data.This study explores online training using temporal-preserving latent variable models.Within online training,there are two main methods:one focuses on reconstructing data and the other on interpreting the data components.Both are considered to evaluate how they diagnose faults over time.Using two experimental datasets,the study confirms that models from both training modes can detect changes in machinery health and identify faults even under varying conditions.Importantly,the complementarity of offline and online models is emphasized,reassuring their versatility in fault diagnostics.Understanding the implications of the training approach and the available model formulations is crucial for further research in latent variable modelbased fault diagnostics.展开更多
Stop frequency models, as one of the elements of activity based models, represent an important part of travel behavior. Unobserved heterogeneity across the travelers should be taken into consideration to prevent biase...Stop frequency models, as one of the elements of activity based models, represent an important part of travel behavior. Unobserved heterogeneity across the travelers should be taken into consideration to prevent biasedness and inconsistency in the estimated parameters in the stop frequency models. Additionally, previous studies on the stop frequency have mostly been done in larger metropolitan areas and less attention has been paid to the areas with less population. This study addresses these gaps by using 2012 travel data from a medium sized U.S. urban area using the work tour for the case study. Stop in the work tour were classified into three groups of outbound leg, work based subtour, and inbound leg of the commutes. Latent Class Poisson Regression Models were used to analyze the data. The results indicate the presence of heterogeneity across the commuters. Using latent class models significantly improves the predictive power of the models compared to regular one class Poisson regression models. In contrast to one class Poisson models, gender becomes insignificant in predicting the number of tours when unobserved heterogeneity is accounted for. The commuters are associated with increased stops on their work based subtour when the employment density of service-related occupations increases in their work zone, but employment density of retail employment does not significantly contribute to the stop making likelihood of the commuters. Additionally, an increase in the number of work tours was associated with fewer stops on the inbound leg of the commute. The results of this study suggest the consideration of unobserved heterogeneity in the stop frequency models and help transportation agencies and policy makers make better inferences from such models.展开更多
A growing stream of study stresses the relevance of subjective elements in understanding the hierarchy of preferences that underpin individual travel behavior. The purpose of this study is to evaluate the impact of va...A growing stream of study stresses the relevance of subjective elements in understanding the hierarchy of preferences that underpin individual travel behavior. The purpose of this study is to evaluate the impact of various factors on mode choice. To achieve this, a multinomial logit model (MNL) was used to analyze the relationships between mode choice and three classes of attributes;Combined Active and Latent, Active only and Latent only attributes. The data used are derived from surveys in the port city of Douala, Cameroon as a case study. Results stipulated that, the combined attributes model performed better than both active only attributes and latent only attributes models. Likewise, latent only attributes model performed better than active only attributes model. The advantage of modelling all three groups is for better selection of the most relevant attributes, and this is very relevant in understanding travel behavior of individuals and mode choice decisions.展开更多
This paper discusses the utilization of latent variable modeling related to occupational health and safety in the mining industry.Latent variable modeling,which is a statistical model that relates observable and laten...This paper discusses the utilization of latent variable modeling related to occupational health and safety in the mining industry.Latent variable modeling,which is a statistical model that relates observable and latent variables,could be used to facilitate researchers’understandings of the underlying constructs or hypothetical factors and their magnitude of effect that constitute a complex system.This enhanced understanding,in turn,can help emphasize the important factors to improve mine safety.The most commonly used techniques include the exploratory factor analysis(EFA),the confirmatory factor analysis(CFA)and the structural equation model with latent variables(SEM).A critical comparison of the three techniques regarding mine safety is provided.Possible applications of latent variable modeling in mining engineering are explored.In this scope,relevant research papers were reviewed.They suggest that the application of such methods could prove useful in mine accident and safety research.Application of latent variables analysis in cognitive work analysis was proposed to improve the understanding of human-work relationships in mining operations.展开更多
High-dimensional and sparse(HiDS)matrices commonly arise in various industrial applications,e.g.,recommender systems(RSs),social networks,and wireless sensor networks.Since they contain rich information,how to accurat...High-dimensional and sparse(HiDS)matrices commonly arise in various industrial applications,e.g.,recommender systems(RSs),social networks,and wireless sensor networks.Since they contain rich information,how to accurately represent them is of great significance.A latent factor(LF)model is one of the most popular and successful ways to address this issue.Current LF models mostly adopt L2-norm-oriented Loss to represent an HiDS matrix,i.e.,they sum the errors between observed data and predicted ones with L2-norm.Yet L2-norm is sensitive to outlier data.Unfortunately,outlier data usually exist in such matrices.For example,an HiDS matrix from RSs commonly contains many outlier ratings due to some heedless/malicious users.To address this issue,this work proposes a smooth L1-norm-oriented latent factor(SL-LF)model.Its main idea is to adopt smooth L1-norm rather than L2-norm to form its Loss,making it have both strong robustness and high accuracy in predicting the missing data of an HiDS matrix.Experimental results on eight HiDS matrices generated by industrial applications verify that the proposed SL-LF model not only is robust to the outlier data but also has significantly higher prediction accuracy than state-of-the-art models when they are used to predict the missing data of HiDS matrices.展开更多
基金supported in part by the National Natural Science Foundation of China (6177249391646114)+1 种基金Chongqing research program of technology innovation and application (cstc2017rgzn-zdyfX0020)in part by the Pioneer Hundred Talents Program of Chinese Academy of Sciences
文摘Latent factor(LF)models are highly effective in extracting useful knowledge from High-Dimensional and Sparse(HiDS)matrices which are commonly seen in various industrial applications.An LF model usually adopts iterative optimizers,which may consume many iterations to achieve a local optima,resulting in considerable time cost.Hence,determining how to accelerate the training process for LF models has become a significant issue.To address this,this work proposes a randomized latent factor(RLF)model.It incorporates the principle of randomized learning techniques from neural networks into the LF analysis of HiDS matrices,thereby greatly alleviating computational burden.It also extends a standard learning process for randomized neural networks in context of LF analysis to make the resulting model represent an HiDS matrix correctly.Experimental results on three HiDS matrices from industrial applications demonstrate that compared with state-of-the-art LF models,RLF is able to achieve significantly higher computational efficiency and comparable prediction accuracy for missing data.I provides an important alternative approach to LF analysis of HiDS matrices,which is especially desired for industrial applications demanding highly efficient models.
基金Supported by the Beijing University of Traditional Chinese Medicine Foundation(No.2015-JYB-JSMS099)the National Science Foundation of China(No.81473527)
文摘OBJECTIVE: To explore the features of Traditional Chinese Medicine(TCM) syndromes in male infertility using computer-based analyses.METHODS: Latent class analysis was used to analyze the TCM syndrome data from 813 patients with male infertility and establish a latent tree model.RESULTS: A latent tree model with a Bayesian information criterion score of-11 263 was created.This model revealed that the characteristics of basic TCM syndromes in patients with male infertility were kidney Yang deficiency, kidney Qi deficiency,spleen Yang deficiency, liver Qi stagnation, Qi stagnation and blood stasis, and dump-heat; moreover,most patients with male infertility had complex syndromes(spleen-kidney Yang deficiency and liver Qi stagnation) rather than simple single syndromes.CONCLUSION: The hidden tree model analysis revealed the objective and quantitative complex relationships between the TCM symptoms of male infertility, and obtained the quantification and objective evidence of TCM syndromes in male infertility.
文摘Mining stimulates environmental and economic impacts on the neighboring community right from the inception to the closure of its operations. The society in the neighborhood of mining gradually adopts a characteristic life-style that is highly influenced by the mining. In order to sustain the societal development beyond the mine closure, it is necessary to plan post mining activities in the area. Thus, it is essential to predict the impacts of mine closure well before the closure. Many societal and family attributes are affected by mine closure. Impact on these attributes is reflected on the overall quality of life of the neighboring community. There are no adequate indicators and/or methodology available to measure social impacts of mine closure on a neighboring community. This paper made an attempt to develop such methodology to predict the degree of adverse effects of mine closure on the quality of life of neighboring communities using the Structural Equation Modeling (SEM) and the Latent Variables Interaction Model (LVM).
文摘Latent class analysis (LCA) is a widely used statistical technique for identifying subgroups in the population based upon multiple indicator variables. It has a number of advantages over other unsupervised grouping procedures such as cluster analysis, including stronger theoretical underpinnings, more clearly defined measures of model fit, and the ability to conduct confirmatory analyses. In addition, it is possible to ascertain whether an LCA solution is equally applicable to multiple known groups, using invariance assessment techniques. This study compared the effectiveness of multiple statistics for detecting group LCA invariance, including a chi-square difference test, a bootstrap likelihood ratio test, and several information indices. Results of the simulation study found that the bootstrap likelihood ratio test was the optimal invariance assessment statistic. In addition to the simulation, LCA group invariance assessment was demonstrated in an application with the Youth Risk Behavior Survey (YRBS). Implications of the simulation results for practice are discussed.
基金supported by the National Basic Research Program of China(Grant No.2012CB417403)the Strategic Priority Research Program of the Chinese Academy of Sciences(Grant No.XDA05090402)the Opening Project of Key Laboratory of Meteorological Disaster of Ministry of Education of Nanjing University of Information Science and Technology(Grant No.KLME1401)
文摘The climatological mean state, seasonal variation and long-term upward trend of 1979-2005 latent heat flux （LHF） in historical runs of 14 coupled general circulation models from CMIP5 （Coupled Model Intercomparison Project Phase 5） are evaluated against OAFlux （Objectively Analyzed air-sea Fluxes） data. Inter-model diversity of these models in simulating the annual mean climatological LHF is discussed. Results show that the models can capture the climatological LHF fairly well, but the amplitudes are generally overestimated. Model-simulated seasonal variations of LHF match well with observations with overestimated amplitudes. The possible origins of these biases are wind speed biases in the CMIP5 models. Inter-model diversity analysis shows that the overall stronger or weaker LHF over the tropical and subtropical Pacific region, and the meridional variability of LHF, are the two most notable diversities of the CMIP5 models. Regression analysis indicates that the inter-model diversity may come from the diversity of simulated SST and near-surface atmospheric specific humidity. Comparing the observed long-term upward trend, the trends of LHF and wind speed are largely underestimated, while trends of SST and air specific humidity are grossly overestimated, which may be the origins of the model biases in reproducing the trend of LHF.
文摘Background: Workplace violence (WV) towards psychiatric staff has commonly been associated with Posttraumatic Stress Disorder (PTSD). However, prospective studies have shown that not all psychiatric staff who experience workplace violence experience post-traumatic stress. Purpose: We want to examine the longitudinal trajectories of PTSD in this population to identify possible subgroups that might be more at risk. Furthermore, we need to investigate whether certain risk factors of PTSD might identify membership in the subgroups. Method: In a sample of psychiatric staff from 18 psychiatric wards in Denmark who had reported an incident of WV, we used Latent Growth Mixture Modelling (LGMM) and further logistic regression analysis to investigate this. Results: We found three separate PTSD trajectories: a recovering, a delayed-onset, and a moderate-stable trajectory. Higher social support and negative cognitive appraisals about oneself, the world and self-blame predicted membership in the delayed-onset trajectory, while higher social support and lower accept coping predicted membership in the delayed-onset trajectory. Conclusion: Although most psychiatric staff go through a natural recovery, it is important to be aware of and identify staff members who might be struggling long-term. More focus on the factors that might predict these groups should be an important task for psychiatric departments to prevent posttraumatic symptomatology from work.
文摘In this project, we consider obtaining Fourier features via more efficient sampling schemes to approximate the kernel in LFMs. A latent force model (LFM) is a Gaussian process whose covariance functions follow an Exponentiated Quadratic (EQ) form, and the solutions for the cross-covariance are expensive due to the computational complexity. To reduce the complexity of mathematical expressions, random Fourier features (RFF) are applied to approximate the EQ kernel. Usually, the random Fourier features are implemented with Monte Carlo sampling, but this project proposes replacing the Monte-Carlo method with the Quasi-Monte Carlo (QMC) method. The first-order and second-order models’ experiment results demonstrate the decrease in NLPD and NMSE, which revealed that the models with QMC approximation have better performance.
文摘Latent variable models can effectively determine the condition of essential rotating machinery without needing labeled data.These models analyze vibration data via an unsupervised learning strategy.Temporal preservation is necessary to obtain an informative latent manifold for the fault diagnosis task.In a temporalpreserving context,two approaches exist to develop a condition-monitoring methodology:offline and online.For latent variable models,the available training modes are not different.While many traditional methods use offline training,online training can dynamically adjust the latent manifold,possibly leading to better fault signature extraction from the vibration data.This study explores online training using temporal-preserving latent variable models.Within online training,there are two main methods:one focuses on reconstructing data and the other on interpreting the data components.Both are considered to evaluate how they diagnose faults over time.Using two experimental datasets,the study confirms that models from both training modes can detect changes in machinery health and identify faults even under varying conditions.Importantly,the complementarity of offline and online models is emphasized,reassuring their versatility in fault diagnostics.Understanding the implications of the training approach and the available model formulations is crucial for further research in latent variable modelbased fault diagnostics.
文摘Stop frequency models, as one of the elements of activity based models, represent an important part of travel behavior. Unobserved heterogeneity across the travelers should be taken into consideration to prevent biasedness and inconsistency in the estimated parameters in the stop frequency models. Additionally, previous studies on the stop frequency have mostly been done in larger metropolitan areas and less attention has been paid to the areas with less population. This study addresses these gaps by using 2012 travel data from a medium sized U.S. urban area using the work tour for the case study. Stop in the work tour were classified into three groups of outbound leg, work based subtour, and inbound leg of the commutes. Latent Class Poisson Regression Models were used to analyze the data. The results indicate the presence of heterogeneity across the commuters. Using latent class models significantly improves the predictive power of the models compared to regular one class Poisson regression models. In contrast to one class Poisson models, gender becomes insignificant in predicting the number of tours when unobserved heterogeneity is accounted for. The commuters are associated with increased stops on their work based subtour when the employment density of service-related occupations increases in their work zone, but employment density of retail employment does not significantly contribute to the stop making likelihood of the commuters. Additionally, an increase in the number of work tours was associated with fewer stops on the inbound leg of the commute. The results of this study suggest the consideration of unobserved heterogeneity in the stop frequency models and help transportation agencies and policy makers make better inferences from such models.
文摘A growing stream of study stresses the relevance of subjective elements in understanding the hierarchy of preferences that underpin individual travel behavior. The purpose of this study is to evaluate the impact of various factors on mode choice. To achieve this, a multinomial logit model (MNL) was used to analyze the relationships between mode choice and three classes of attributes;Combined Active and Latent, Active only and Latent only attributes. The data used are derived from surveys in the port city of Douala, Cameroon as a case study. Results stipulated that, the combined attributes model performed better than both active only attributes and latent only attributes models. Likewise, latent only attributes model performed better than active only attributes model. The advantage of modelling all three groups is for better selection of the most relevant attributes, and this is very relevant in understanding travel behavior of individuals and mode choice decisions.
基金Natural Sciences and Engineering Research Council of Canada(NSERC)(ID:236482)for supporting this research
文摘This paper discusses the utilization of latent variable modeling related to occupational health and safety in the mining industry.Latent variable modeling,which is a statistical model that relates observable and latent variables,could be used to facilitate researchers’understandings of the underlying constructs or hypothetical factors and their magnitude of effect that constitute a complex system.This enhanced understanding,in turn,can help emphasize the important factors to improve mine safety.The most commonly used techniques include the exploratory factor analysis(EFA),the confirmatory factor analysis(CFA)and the structural equation model with latent variables(SEM).A critical comparison of the three techniques regarding mine safety is provided.Possible applications of latent variable modeling in mining engineering are explored.In this scope,relevant research papers were reviewed.They suggest that the application of such methods could prove useful in mine accident and safety research.Application of latent variables analysis in cognitive work analysis was proposed to improve the understanding of human-work relationships in mining operations.
基金supported in part by the National Natural Science Foundation of China(61702475,61772493,61902370,62002337)in part by the Natural Science Foundation of Chongqing,China(cstc2019jcyj-msxmX0578,cstc2019jcyjjqX0013)+1 种基金in part by the Chinese Academy of Sciences“Light of West China”Program,in part by the Pioneer Hundred Talents Program of Chinese Academy of Sciencesby Technology Innovation and Application Development Project of Chongqing,China(cstc2019jscx-fxydX0027)。
文摘High-dimensional and sparse(HiDS)matrices commonly arise in various industrial applications,e.g.,recommender systems(RSs),social networks,and wireless sensor networks.Since they contain rich information,how to accurately represent them is of great significance.A latent factor(LF)model is one of the most popular and successful ways to address this issue.Current LF models mostly adopt L2-norm-oriented Loss to represent an HiDS matrix,i.e.,they sum the errors between observed data and predicted ones with L2-norm.Yet L2-norm is sensitive to outlier data.Unfortunately,outlier data usually exist in such matrices.For example,an HiDS matrix from RSs commonly contains many outlier ratings due to some heedless/malicious users.To address this issue,this work proposes a smooth L1-norm-oriented latent factor(SL-LF)model.Its main idea is to adopt smooth L1-norm rather than L2-norm to form its Loss,making it have both strong robustness and high accuracy in predicting the missing data of an HiDS matrix.Experimental results on eight HiDS matrices generated by industrial applications verify that the proposed SL-LF model not only is robust to the outlier data but also has significantly higher prediction accuracy than state-of-the-art models when they are used to predict the missing data of HiDS matrices.
