Search results for: Cox regression model

Authors: Mitra Sadoughi

Abstract:

This study presents the knowledge-sharing model based on individual and organizational factors related to faculty members. To achieve this goal, individual and organizational factors were presented through qualitative research in the form of open codes, axial, and selective observations; then, the final model was obtained using structural equation model. Participants included 1,719 faculty members of the Azad Universities, Mazandaran Province, Region 3. The samples related to the qualitative survey included 25 faculty members experienced at teaching and the samples related to the quantitative survey included 326 faculty members selected by multistage cluster sampling. A 72-item questionnaire was used to measure the quantitative variables. The reliability of the questionnaire was 0.93. Its content and face validity was determined with the help of faculty members, consultants, and other experts. For the analysis of quantitative data obtained from structural model and regression, SPSS and LISREL were used. The results showed that the status of knowledge sharing is moderate in the universities. Individual factors influencing knowledge sharing included the sharing of educational materials, perception, confidence and knowledge self-efficiency, and organizational factors influencing knowledge sharing included structural social capital, cognitive social capital, social capital relations, organizational communication, organizational structure, organizational culture, IT infrastructure and systems of rewards. Finally, it was found that the contribution of individual factors on knowledge sharing was more than organizational factors; therefore, a model was presented in which contribution of individual and organizational factors were determined.

Keywords: knowledge sharing, social capital, organizational communication, knowledge self-efficiency, perception, trust, organizational culture

Procedia PDF Downloads 392

Authors: Qasem Qananwah, Ahmad Dagamseh, Hiam AlQuran, Khalid Shaker Ibrahim

Abstract:

Blood pressure helps the physicians greatly to have a deep insight into the cardiovascular system. The determination of individual blood pressure is a standard clinical procedure considered for cardiovascular system problems. The conventional techniques to measure blood pressure (e.g. cuff method) allows a limited number of readings for a certain period (e.g. every 5-10 minutes). Additionally, these systems cause turbulence to blood flow; impeding continuous blood pressure monitoring, especially in emergency cases or critically ill persons. In this paper, the most important statistical features in the photoplethysmogram (PPG) signals were extracted to estimate the blood pressure noninvasively. PPG signals from more than 40 subjects were measured and analyzed and 12 features were extracted. The features were fed to principal component analysis (PCA) to find the most important independent features that have the highest correlation with blood pressure. The results show that the stiffness index means and standard deviation for the beat-to-beat heart rate were the most important features. A model representing both features for Systolic Blood Pressure (SBP) and Diastolic Blood Pressure (DBP) was obtained using a statistical regression technique. Surface fitting is used to best fit the series of data and the results show that the error value in estimating the SBP is 4.95% and in estimating the DBP is 3.99%.

Keywords: blood pressure, noninvasive optical system, principal component analysis, PCA, continuous monitoring

Procedia PDF Downloads 161

Authors: Ayman M. Othman, Hassan Y. Ahmed, Tallat A. Ali

Abstract:

Inadequate maintenance programs and funds allocated for highway networks in the developed countries have led to fast deterioration of road side elements. Therefore, this research focuses on developing a performance model for road side elements periodic maintenance activities. Road side elements that receive periodic maintenance include; earthen shoulder, road signs and traffic markings. Using the level of service concept, the developed model can determine the optimal periodic maintenance intervals for those elements based on a selected level of service suitable with the available periodic maintenance budget. Data related to time periods for progressive deterioration stages for the chosen elements were collected. Ten maintenance experts in Aswan, Sohag and Assiut cities were interviewed for that purpose. Time in months related to 10%, 25%, 40%, 50%, 75%, 90% and 100% deterioration of each road side element was estimated based on the experts opinion. Least square regression analysis has shown that a power function represents the best fit for earthen shoulders edge drop-off and damage of road signs with time. It was also evident that, the progressive dirtiness of road signs could be represented by a quadratic function an a linear function could represent the paint degradation nature of both traffic markings and road signs. Actual measurements of earthen shoulder edge drop-off agree considerably with the developed model.

Keywords: deterioration, level of service, periodic maintenance, performance model, road side element

Procedia PDF Downloads 572

Authors: Ubaidillah Zuhdi, Shunsuke Mori, Kazuhisa Kamegai

Abstract:

The purpose of this study is to forecast the influences of Information and Communication Technology (ICT) on the structural changes of Japanese economies based on Leontief Input-Output (IO) coefficients. This study establishes a statistical analysis to predict the future interrelationships among industries. We employ the Constrained Multivariate Regression (CMR) model to analyze the historical changes of input-output coefficients. Statistical significance of the model is then tested by Likelihood Ratio Test (LRT). In our model, ICT is represented by two explanatory variables, i.e. computers (including main parts and accessories) and telecommunications equipment. A previous study, which analyzed the influences of these variables on the structural changes of Japanese industrial sectors from 1985-2005, concluded that these variables had significant influences on the changes in the business circumstances of Japanese commerce, business services and office supplies, and personal services sectors. The projected future Japanese economic structure based on the above forecast generates the differentiated direct and indirect outcomes of ICT penetration.

Keywords: forecast, ICT, industrial structural changes, statistical analysis

Procedia PDF Downloads 375

Authors: Gang Wang

Abstract:

Education is widely believed to be a prerequisite for democracy and civil society, but the causal link between education and outcome variables is usually hardly to be identified. This study applies a fuzzy regression discontinuity design to examine the effects of college education on democratic attitudes in the Chinese context. In the analysis treatment assignment is determined by students’ college entry years and thus naturally selected by subjects’ ages. Using a sample of Chinese college students collected in Beijing in 2009, this study finds that college education actually reduces undergraduates’ motivation for political development in China but promotes political loyalty to the authoritarian government. Further hypotheses tests explain these interesting findings from two perspectives. The first is related to the complexity of politics. As college students progress over time, they increasingly realize the complexity of political reform in China’s authoritarian regime and rather stay away from politics. The second is related to students’ career opportunities. As students are close to graduation, they are immersed with job hunting and have a reduced interest in political freedom.

Keywords: china, college education, democratic attitudes, regression discontinuity

Procedia PDF Downloads 351

Authors: Kamaludeen Samaila, Khashyaullah Abdulfattah, Fahimi Ahmad Bin Amir

Abstract:

The study was conducted to examine the relationship between selected factors (performance expectancy, effort expectancy, social influence and facilitating condition) and students’ intention to use the learning management system (LMS), as well as investigating the factors predicting students’ intention to use the LMS. The study was specifically conducted at the Faculty of Educational Study of University Putra Malaysia. Questionnaires were distributed to 277 respondents using a random sampling technique. SPSS Version 22 was employed in analyzing the data; the findings of this study indicated that performance expectancy (r = .69, p < .01), effort expectancy (r=.60, p < .01), social influence (r = .61, p < .01), and facilitating condition (r=.42, p < .01), were significantly related to students’ intention to use the LMS. In addition, the result also revealed that performance expectancy (β = .436, p < .05), social influence (β=.232, p < .05), and effort expectancy (β = .193, p < .05) were strong predictors of students’ intention to use the LMS. The analysis further indicated that (R2) is 0.054 which means that 54% of variation in the dependent variable is explained by the entire predictor variables entered into the regression model. Understanding the factors that affect students’ intention to use the LMS could help the lecturers, LMS managers and university management to develop the policies that may attract students to use the LMS.

Keywords: LMS, postgraduate students, PutraBlas, students’ intention, UPM, UTAUT model

Procedia PDF Downloads 510

Authors: Muhammad Waseem Akram

Abstract:

The current study attempts to analyze the impact of business constraints on the growth potential and performance of Small and Medium Enterprises (SMEs) in the electrical industry of Pakistan. Primary data have been utilized for the study collected from the electrical industry cluster in Sargodha, Pakistan. OLS regression is used to assess the impact of business constraints on the performance of SMEs by controlling the effect of Technology Level, Innovations, and Firm Size. To associate business constraints with the growth potential of SMEs, the study utilized Tetrachoric Correlation and Logistic Regression. Findings reveal that all the business constraints negatively affect the performance of SMEs in the electrical industry except Political Instability. Results of Tetrachoric Correlation show that all the business constraints are negatively correlated with the growth potential of SMEs. Logistic Regression results show that Energy Constraint, Inflation and Price Instability, and Bad Business Practices, all three business constraints cause to reduce the probability of income growth in sample SMEs.

Keywords: SMEs, business constraints, performance, growth potential

Procedia PDF Downloads 169

Authors: Sifriyani Sifriyani, I Nyoman Budiantara, Sri Haryatmi, Gunardi Gunardi

Abstract:

East Java Province has a first rank as a province that has the most counties and cities in Indonesia and has the largest population. In 2015, the population reached 38.847.561 million, this figure showed a very high population growth. High population growth is feared to lead to increase the levels of unemployment. In this study, the researchers mapped and modeled the unemployment rate with 6 variables that were supposed to influence. Modeling was done by nonparametric geographically weighted regression methods with truncated spline approach. This method was chosen because spline method is a flexible method, these models tend to look for its own estimation. In this modeling, there were point knots, the point that showed the changes of data. The selection of the optimum point knots was done by selecting the most minimun value of Generalized Cross Validation (GCV). Based on the research, 6 variables were declared to affect the level of unemployment in eastern Java. They were the percentage of population that is educated above high school, the rate of economic growth, the population density, the investment ratio of total labor force, the regional minimum wage and the ratio of the number of big industry and medium scale industry from the work force. The nonparametric geographically weighted regression models with truncated spline approach had a coefficient of determination 98.95% and the value of MSE equal to 0.0047.

Keywords: East Java, nonparametric geographically weighted regression, spatial, spline approach, unemployed rate

Procedia PDF Downloads 321

Authors: L. Madouni, M. Ould Ouali, N. E. Hannachi

Abstract:

Two constitutive models for concrete are available in ABAQUS/Explicit, the Brittle Cracking Model and the Concrete Damaged Plasticity Model, and their suitability and limitations are well known. The aim of the present paper is to implement a damage-friction concrete constitutive model and to evaluate the performance of this model by comparing the predicted response with experimental data. The constitutive formulation of this material model is reviewed. In order to have consistent results, the parameter identification and calibration for the model have been performed. Several numerical simulations are presented in this paper, whose results allow for validating the capability of the proposed model for reproducing the typical nonlinear performances of concrete structures under different monotonic and cyclic load conditions. The results of the evaluation will be used for recommendations concerning the application and further improvements of the investigated model.

Keywords: Abaqus, concrete, constitutive model, numerical simulation

Procedia PDF Downloads 365

Authors: Shuo Tian, Xuepiao Bai, Jianqin Shang, Pengtao Gai, Yuansong Zeng

Abstract:

Shot peen forming is an essential forming process for aircraft metal wing panel. With the development of computer simulation technology, scholars have proposed a numerical simulation method of shot peen forming based on stress field. Three shot peen forming indexes of crater diameter, shot speed and surface coverage are required as simulation parameters in the stress field method. It is necessary to establish the relationship between simulation and experimental process parameters in order to simulate the deformation under different shot peen forming parameters. The shot peen forming tests of the 2024-T351 aluminum alloy workpieces were carried out using uniform test design method, and three factors of air pressure, feed rate and shot flow were selected. The second-order response surface model between simulation parameters and uniform test factors was established by stepwise regression method using MATLAB software according to the results. The response surface model was combined with the stress field method to simulate the shot peen forming deformation of the workpiece. Compared with the experimental results, the simulated values were smaller than the corresponding test values, the maximum and average errors were 14.8% and 9%, respectively.

Keywords: shot peen forming, process parameter, response surface model, numerical simulation

Procedia PDF Downloads 87

Authors: Haolin Li, Eric Bair, Jane Monaco, Quefeng Li

Abstract:

The temporomandibular disorder (TMD) is a series of musculoskeletal disorders ranging from jaw pain to chronic debilitating pain, and the risk factors for the onset and maintenance of TMD are still unclear. Prior researches have shown that the potential risk factors for chronic TMD are related to psychosocial factors, autonomic functions, and pain sensitivity. Using data from the Orofacial Pain: Prospective Evaluation and Risk Assessment (OPPERA) study’s baseline case-control study, we examine whether the risk factors identified by prior researches are still statistically significant after taking all of the risk measures into account in one single model, and we also compare the relative influences of the risk factors in three different perspectives (psychosocial factors, autonomic functions, and pain sensitivity) on the chronic TMD. The statistical analysis is conducted by using ridge logistic regression and bootstrapping, in which the performance of the algorithms has been assessed using extensive simulation studies. The results support most of the findings of prior researches that there are many psychosocial and pain sensitivity measures that have significant associations with chronic TMD. However, it is surprising that most of the risk factors of autonomic functions have not presented significant associations with chronic TMD, as described by a prior research.

Keywords: autonomic function, OPPERA study, pain sensitivity, psychosocial measures, temporomandibular disorder

Procedia PDF Downloads 188

Authors: Morteza Raei Dehaghi, Seyed Mohammad Mirhashemi

Abstract:

The current study explored the effect of economic development, financial development and institutional quality on environmental destruction in upper-middle income countries during the time period of 1999-2011. The dependent variable is logarithm of carbon dioxide emissions that can be considered as an index for destruction or quality of the environment given to its effects on the environment. Financial development and institutional development variables as well as some control variables were considered. In order to study cross-sectional correlation among the countries under study, Pesaran and Friz test was used. Since the results of both tests show cross-sectional correlation in the countries under study, seemingly unrelated regression method was utilized for model estimation. The results disclosed that Kuznets’ environmental curve hypothesis is confirmed in upper-middle income countries and also, financial development and institutional quality have a significant effect on environmental quality. The results of this study can be considered by policy makers in countries with different income groups to have access to a growth accompanied by improved environmental quality.

Keywords: economic development, environmental destruction, financial development, institutional development, seemingly unrelated regression

Procedia PDF Downloads 348

Authors: Jiahao Tian, Michael D. Porter

Abstract:

Law enforcement agencies are tasked with crime prevention and crime reduction under limited resources. Having an accurate temporal estimate of the crime rate would be valuable to achieve such a goal. However, estimation is usually complicated by the interval-censored nature of crime data. We cast the problem of intensity estimation as a Poisson regression using an EM algorithm to estimate the parameters. Two special penalties are added that provide smoothness over the time of day and day of the week. This approach presented here provides accurate intensity estimates and can also uncover day-of-week clusters that share the same intensity patterns. Anticipating where and when crimes might occur is a key element to successful policing strategies. However, this task is complicated by the presence of interval-censored data. The censored data refers to the type of data that the event time is only known to lie within an interval instead of being observed exactly. This type of data is prevailing in the field of criminology because of the absence of victims for certain types of crime. Despite its importance, the research in temporal analysis of crime has lagged behind the spatial component. Inspired by the success of solving crime-related problems with a statistical approach, we propose a statistical model for the temporal intensity estimation of crime with censored data. The model is built on Poisson regression and has special penalty terms added to the likelihood. An EM algorithm was derived to obtain maximum likelihood estimates, and the resulting model shows superior performance to the competing model. Our research is in line with the smart policing initiative (SPI) proposed by the Bureau Justice of Assistance (BJA) as an effort to support law enforcement agencies in building evidence-based, data-driven law enforcement tactics. The goal is to identify strategic approaches that are effective in crime prevention and reduction. In our case, we allow agencies to deploy their resources for a relatively short period of time to achieve the maximum level of crime reduction. By analyzing a particular area within cities where data are available, our proposed approach could not only provide an accurate estimate of intensities for the time unit considered but a time-variation crime incidence pattern. Both will be helpful in the allocation of limited resources by either improving the existing patrol plan with the understanding of the discovery of the day of week cluster or supporting extra resources available.

Keywords: cluster detection, EM algorithm, interval censoring, intensity estimation

Procedia PDF Downloads 66

Authors: Paul R Armstrong

Abstract:

Doubled haploids (DHs) have become an important breeding tool for creating maize inbred lines, although several bottlenecks in the DH production process limit wider development, application, and adoption of the technique. DH kernels are typically sorted manually and represent about 10% of the seeds in a much larger pool where the remaining 90% are hybrid siblings. This introduces time constraints on DH production and manual sorting is often not accurate. Automated sorting based on the chemical composition of the kernel can be effective, but devices, namely NMR, have not achieved the sorting speed to be a cost-effective replacement to manual sorting. This study evaluated a single kernel near-infrared reflectance spectroscopy (skNIR) platform to accurately identify DH kernels based on oil content. The skNIR platform is a higher-throughput device, approximately 3 seeds/s, that uses spectra to predict oil content of each kernel from maize crosses intentionally developed to create larger than normal oil differences, 1.5%-2%, between DH and hybrid kernels. Spectra from the skNIR were used to construct a partial least squares regression (PLS) model for oil and for a categorical reference model of 1 (DH kernel) or 2 (hybrid kernel) and then used to sort several crosses to evaluate performance. Two approaches were used for sorting. The first used a general PLS model developed from all crosses to predict oil content and then used for sorting each induction cross, the second was the development of a specific model from a single induction cross where approximately fifty DH and one hundred hybrid kernels used. This second approach used a categorical reference value of 1 and 2, instead of oil content, for the PLS model and kernels selected for the calibration set were manually referenced based on traditional commercial methods using coloration of the tip cap and germ areas. The generalized PLS oil model statistics were R2 = 0.94 and RMSE = .93% for kernels spanning an oil content of 2.7% to 19.3%. Sorting by this model resulted in extracting 55% to 85% of haploid kernels from the four induction crosses. Using the second method of generating a model for each cross yielded model statistics ranging from R2s = 0.96 to 0.98 and RMSEs from 0.08 to 0.10. Sorting in this case resulted in 100% correct classification but required models that were cross. In summary, the first generalized model oil method could be used to sort a significant number of kernels from a kernel pool but was not close to the accuracy of developing a sorting model from a single cross. The penalty for the second method is that a PLS model would need to be developed for each individual cross. In conclusion both methods could find useful application in the sorting of DH from hybrid kernels.

Keywords: NIR, haploids, maize, sorting

Procedia PDF Downloads 302

Authors: Umar Hayatu Sidik

Abstract:

The natural gas vehicle represents a cost-competitive, lower-emission alternative to the gasoline-fuelled vehicle. The immediate challenge that confronts natural gas is increasing its energy density. This paper addresses the question of energy density by reviewing the storage technologies for natural gas with improved adsorbent. Technical comparisons are made between storage systems containing adsorbent and conventional compressed natural gas based on the associated amount of moles contained with Compressed Natural Gas (CNG) and Adsorbed Natural Gas (ANG). We also compare gas storage in different cylinder types (1, 2, 3 and 4) based on weight factor and storage capacity. For the storage tank system, we discussed the concept of carbon adsorbents, when used in CNG tanks, offer a means of increasing onboard fuel storage and, thereby, increase the driving range of the vehicle. It confirms that the density of the stored gas in ANG is higher than that of compressed natural gas (CNG) operated at the same pressure. The obtained experimental data were correlated using linear regression analysis with common adsorption kinetic (Pseudo-first order and Pseudo-second order) and isotherm models (Sip and Toth). The pseudo-second-order kinetics describe the best fitness with a correlation coefficient of 9945 at 35 bar. For adsorption isotherms, the Sip model shows better fitness with the regression coefficient (R2) of 0.9982 and with the lowest RSMD value of 0.0148. The findings revealed the potential of adsorbent in natural gas storage applications.

Keywords: natural gas, adsorbent, compressed natural gas, adsorption

Procedia PDF Downloads 60

Authors: Puttharapong Sakulwaropas, Uraiwan Jaroengeratikun

Abstract:

Casualty insurance business, the optimal premium pricing and adequate cost for an insurance company are important in risk management. Normally, the insurance pure premium can be determined by multiplying the claim frequency with the claim cost. The aim of this research was to study in the application of generalized linear models to select the risk factor for model of claim frequency and claim cost for estimating a pure premium. In this study, the data set was the claim of comprehensive motor insurance, which was provided by one of the insurance company in Thailand. The results of this study found that the risk factors significantly related to pure premium at the 0.05 level consisted of no claim bonus (NCB) and used of the car (Car code).

Keywords: generalized linear models, risk factor, pure premium, regression model

Procedia PDF Downloads 466

Authors: Mesfin Nigussie

Abstract:

The objectives of the study were to identify factors affecting rural households’ willingness to install biogas plant and amount willingness to pay in order to examine determinants of effective demand for biogas technology. A multistage sampling technique was employed to select 120 respondents for the study. The binary probit regression model was employed to identify factors affecting rural households’ decision to install biogas technology. The probit model result revealed that household size, total household income, access to extension services related to biogas, access to credit service, proximity to water sources, perception of households about the quality of biogas, perception index about attributes of biogas, perception of households about installation cost of biogas and availability of energy source were statistically significant in determining household’s decision to install biogas. Tobit model was employed to examine determinants of rural household’s amount of willingness to pay. Based on the model result, age of the household head, total annual income of the household, access to extension service and availability of other energy source were significant variables that influence willingness to pay. Providing due considerations for extension services, availability of credit or subsidy, improving the quality of biogas technology design and minimizing cost of installation by using locally available materials are the main suggestions of this research that help to create effective demand for biogas technology.

Keywords: biogas technology, effective demand, probit model, tobit model, willingnes to pay

Procedia PDF Downloads 140

Authors: V. Jamuna, A. K. Chakravarty, C. S. Patil, Vijay Kumar, M. A. Mir, Rakesh Kumar

Abstract:

Intense selection of buffaloes for milk production at organized herds of the country without giving due attention to fertility traits viz. pregnancy rate has lead to deterioration in their performances. Aim of study is to develop an optimum model for predicting pregnancy rate and to assess the level of pregnancy rate with respect to milk production Murrah buffaloes. Data pertaining to 1224 lactation records of Murrah buffaloes spread over a period 21 years were analyzed and it was observed that pregnancy rate depicted negative phenotypic association with lactation milk yield (-0.08 ± 0.04). For developing optimum model for pregnancy rate in Murrah buffaloes seven simple and multiple regression models were developed. Among the seven models, model II having only Service period as an independent reproduction variable, was found to be the best prediction model, based on the four statistical criterions (high coefficient of determination (R 2), low mean sum of squares due to error (MSSe), conceptual predictive (CP) value, and Bayesian information criterion (BIC). For standardizing the level of fertility with milk production, pregnancy rate was classified into seven classes with the increment of 10% in all parities, life time and their corresponding average pregnancy rate in relation to the average lactation milk yield (MY).It was observed that to achieve around 2000 kg MY which can be considered optimum for Indian Murrah buffaloes, level of pregnancy rate should be in between 30-50%.

Keywords: life time, pregnancy rate, production, service period, standardization

Procedia PDF Downloads 635

Authors: Arsalan Ghaderi, Behzad Karami Matin, Abdolrahim Afkhamzadeh, Abouzar Keshavarzi, Parvin Nokhasi

Abstract:

Background: Lack of individual health as one of the major health problems among the pilgrims can be followed by several complications. The main aim of this study was to determine factors predicting individual health among pilgrims of Kurdistan County; in the west of Iran and health belief model (HBM) was applied as theoretical framework. Methods: A cross-sectional study was conducted among 100 pilgrims who referred in the red crescent of Kurdistan County, the west of Iran which was randomly selected for participation in this study. A structured questionnaire was applied for collecting data and data were analyzed by SPSS version 21 using bivariate correlations and linear regression statistical tests. Results: The mean age of respondents was 59.45 years [SD: 11.56], ranged from 50 to 73 years. The HBM predictor variables accounted for 47% of the variation in the outcome measure of the individual health. The best predictors for individual health were perceived severity and cause to action. Conclusion: Based on our result, it seems that designing and implementation of educational programs to increase seriousness about complications of lack of individual health and increasing cause to action among the pilgrims may be useful in order to promote individual health among pilgrims.

Keywords: individual health, pilgrims, Iran, health belief model

Procedia PDF Downloads 529

Authors: Girish Chavadappanavar

Abstract:

The impact of climate on agriculture could result in problems with food security and may threaten the livelihood activities upon which much of the population depends. In the present study, the development of a statistical yield forecast model has been carried out for sugarcane production over Belagavi district, Karnataka using weather variables of crop growing season and past observed yield data for the period of 1971 to 2010. The study shows that this type of statistical yield forecast model could efficiently forecast yield 5 weeks and even 10 weeks in advance of the harvest for sugarcane within an acceptable limit of error. The performance of the model in predicting yields at the district level for sugarcane crops is found quite satisfactory for both validation (2007 and 2008) as well as forecasting (2009 and 2010).In addition to the above study, the climate variability of the area has also been studied, and hence, the data series was tested for Mann Kendall Rank Statistical Test. The maximum and minimum temperatures were found to be significant with opposite trends (decreasing trend in maximum and increasing in minimum temperature), while the other three are found in significant with different trends (rainfall and evening time relative humidity with increasing trend and morning time relative humidity with decreasing trend).

Keywords: climate impact, regression analysis, yield and forecast model, sugar models

Procedia PDF Downloads 71

Authors: Debarati Bhaumik, Diptish Dey

Abstract:

Multilevel models, increasingly deployed in industries such as insurance, food production, and entertainment within functions such as marketing and supply chain management, need to be transparent and ethical. Applications usually result in binary classification within groups or hierarchies based on a set of input features. Using open-source datasets, we demonstrate that popular explainability methods, such as SHAP and LIME, consistently underperform inaccuracy when interpreting these models. They fail to predict the order of feature importance, the magnitudes, and occasionally even the nature of the feature contribution (negative versus positive contribution to the outcome). Besides accuracy, the computational intractability of SHAP for binomial classification is a cause of concern. For transparent and ethical applications of these hierarchical statistical models, sound audit frameworks need to be developed. In this paper, we propose an audit framework for technical assessment of multilevel regression models focusing on three aspects: (i) model assumptions & statistical properties, (ii) model transparency using different explainability methods, and (iii) discrimination assessment. To this end, we undertake a quantitative approach and compare intrinsic model methods with SHAP and LIME. The framework comprises a shortlist of KPIs, such as PoCE (Percentage of Correct Explanations) and MDG (Mean Discriminatory Gap) per feature, for each of these three aspects. A traffic light risk assessment method is furthermore coupled to these KPIs. The audit framework will assist regulatory bodies in performing conformity assessments of AI systems using multilevel binomial classification models at businesses. It will also benefit businesses deploying multilevel models to be future-proof and aligned with the European Commission’s proposed Regulation on Artificial Intelligence.

Keywords: audit, multilevel model, model transparency, model explainability, discrimination, ethics

Procedia PDF Downloads 95

Authors: Haifeng Wang, Haili Zhang

Abstract:

Most movie recommendation systems have been developed for customers to find items of interest. This work introduces a predictive model usable by small and medium-sized enterprises (SMEs) who are in need of a data-based and analytical approach to stock proper movies for local audiences and retain more customers. We used classification models to extract features from thousands of customers’ demographic, behavioral and social information to predict their movie genre preference. In the implementation, a Gaussian kernel support vector machine (SVM) classification model and a logistic regression model were established to extract features from sample data and their test error-in-sample were compared. Comparison of error-out-sample was also made under different Vapnik–Chervonenkis (VC) dimensions in the machine learning algorithm to find and prevent overfitting. Gaussian kernel SVM prediction model can correctly predict movie genre preferences in 85% of positive cases. The accuracy of the algorithm increased to 93% with a smaller VC dimension and less overfitting. These findings advance our understanding of how to use machine learning approach to predict customers’ preferences with a small data set and design prediction tools for these enterprises.

Keywords: computational social science, movie preference, machine learning, SVM

Procedia PDF Downloads 260

Authors: Farhad Asadi, Mohammad Javad Mollakazemi, Aref Ghafouri

Abstract:

Recent research in neural networks science and neuroscience for modeling complex time series data and statistical learning has focused mostly on learning from high input space and signals. Local linear models are a strong choice for modeling local nonlinearity in data series. Locally weighted projection regression is a flexible and powerful algorithm for nonlinear approximation in high dimensional signal spaces. In this paper, different learning scenario of one and two dimensional data series with different distributions are investigated for simulation and further noise is inputted to data distribution for making different disordered distribution in time series data and for evaluation of algorithm in locality prediction of nonlinearity. Then, the performance of this algorithm is simulated and also when the distribution of data is high or when the number of data is less the sensitivity of this approach to data distribution and influence of important parameter of local validity in this algorithm with different data distribution is explained.

Keywords: local nonlinear estimation, LWPR algorithm, online training method, locally weighted projection regression method

Procedia PDF Downloads 502

Authors: E. Tchandao Mangamana, V. Cariou, E. Vigneau, R. Glele Kakai, E. M. Qannari

Abstract:

A new definition of a latent variable associated with a dataset makes it possible to propose variants of the PLS2 regression and the multi-block PLS (MB-PLS). We shall refer to these variants as Rd-PLS regression and Rd-MB-PLS respectively because they are inspired by both Redundancy analysis and PLS regression. Usually, a latent variable t associated with a dataset Z is defined as a linear combination of the variables of Z with the constraint that the length of the loading weights vector equals 1. Formally, t=Zw with ‖w‖=1. Denoting by Z' the transpose of Z, we define herein, a latent variable by t=ZZ’q with the constraint that the auxiliary variable q has a norm equal to 1. This new definition of a latent variable entails that, as previously, t is a linear combination of the variables in Z and, in addition, the loading vector w=Z’q is constrained to be a linear combination of the rows of Z. More importantly, t could be interpreted as a kind of projection of the auxiliary variable q onto the space generated by the variables in Z, since it is collinear to the first PLS1 component of q onto Z. Consider the situation in which we aim to predict a dataset Y from another dataset X. These two datasets relate to the same individuals and are assumed to be centered. Let us consider a latent variable u=YY’q to which we associate the variable t= XX’YY’q. Rd-PLS consists in seeking q (and therefore u and t) so that the covariance between t and u is maximum. The solution to this problem is straightforward and consists in setting q to the eigenvector of YY’XX’YY’ associated with the largest eigenvalue. For the determination of higher order components, we deflate X and Y with respect to the latent variable t. Extending Rd-PLS to the context of multi-block data is relatively easy. Starting from a latent variable u=YY’q, we consider its ‘projection’ on the space generated by the variables of each block Xk (k=1, ..., K) namely, tk= XkXk'YY’q. Thereafter, Rd-MB-PLS seeks q in order to maximize the average of the covariances of u with tk (k=1, ..., K). The solution to this problem is given by q, eigenvector of YY’XX’YY’, where X is the dataset obtained by horizontally merging datasets Xk (k=1, ..., K). For the determination of latent variables of order higher than 1, we use a deflation of Y and Xk with respect to the variable t= XX’YY’q. In the same vein, extending Rd-MB-PLS to the path modeling setting is straightforward. Methods are illustrated on the basis of case studies and performance of Rd-PLS and Rd-MB-PLS in terms of prediction is compared to that of PLS2 and MB-PLS.

Keywords: multiblock data analysis, partial least squares regression, path modeling, redundancy analysis

Procedia PDF Downloads 147

Authors: Arslan Murtaza

Abstract:

Model Driven Architecture (MDA) is technique presented by OMG (Object Management Group) for software development in which different models are proposed and converted them into code. The main plan is to identify task by using PIM (Platform Independent Model) and transform it into PSM (Platform Specific Model) and then converted into code. In this review paper describes some challenges and issues that are faced in MDA, type and transformation of models (e.g. CIM, PIM and PSM), and evaluation of MDA-based methodologies.

Keywords: OMG, model driven rrchitecture (MDA), computation independent model (CIM), platform independent model (PIM), platform specific model(PSM), MDA-based methodologies

Procedia PDF Downloads 459

Authors: Zachary Pan

Abstract:

Our general mission is to solve the problem of classifying images into different body part types and deciding if each of them is healthy or not. However, for now, we will determine healthiness for only one-sixth of the body parts, specifically the chest. We will detect pneumonia in X-ray scans of those chest images. With this type of AI, doctors can use it as a second opinion when they are taking CT or X-ray scans of their patients. Another ad-vantage of using this machine learning classifier is that it has no human weaknesses like fatigue. The overall ap-proach to this problem is to split the problem into two parts: first, classify the image, then determine if it is healthy. In order to classify the image into a specific body part class, the body parts dataset must be split into test and training sets. We can then use many models, like neural networks or logistic regression models, and fit them using the training set. Now, using the test set, we can obtain a realistic accuracy the models will have on images in the real world since these testing images have never been seen by the models before. In order to increase this testing accuracy, we can also apply many complex algorithms to the models, like multiplicative weight update. For the second part of the problem, to determine if the body part is healthy, we can have another dataset consisting of healthy and non-healthy images of the specific body part and once again split that into the test and training sets. We then use another neural network to train on those training set images and use the testing set to figure out its accuracy. We will do this process only for the chest images. A major conclusion reached is that convolutional neural networks are the most reliable and accurate at image classification. In classifying the images, the logistic regression model, the neural network, neural networks with multiplicative weight update, neural networks with the black box algorithm, and the convolutional neural network achieved 96.83 percent accuracy, 97.33 percent accuracy, 97.83 percent accuracy, 96.67 percent accuracy, and 98.83 percent accuracy, respectively. On the other hand, the overall accuracy of the model that de-termines if the images are healthy or not is around 78.37 percent accuracy.

Keywords: body part, healthcare, machine learning, neural networks

Procedia PDF Downloads 103

Authors: Katherine Maurer

Abstract:

Background and purpose: Violence that occurs within families is a global social problem. Children who are victims or witness to family violence are at risk for many negative effects both proximally and distally. One of the most disconcerting long-term effects occurs when child victims become adult perpetrators: the intergenerational transmission of family violence (ITFV). Early identification of those children most at risk for ITFV is needed to inform interventions to prevent future family violence perpetration and victimization. Only about 25-30% of child family violence victims become perpetrators of adult family violence (either child abuse, partner abuse, or both). Prior research has primarily been conducted using dichotomous measures of exposure (yes; no) to predict ITFV, given the low incidence rate in community samples. It is often assumed that exposure to greater amounts of violence predicts greater risk of ITFV. However, no previous longitudinal study with a community sample has tested a dose-response model of exposure to physical child abuse and parental physical intimate partner violence (IPV) using count data of frequency and severity of violence to predict adult ITFV. The current study used advanced statistical methods to test if increased childhood exposure would predict greater risk of ITFV. Methods: The study utilized 3 panels of prospective data from a cohort of 15 year olds (N=338) from the Project on Human Development in Chicago Neighborhoods longitudinal study. The data were comprised of a stratified probability sample of seven ethnic/racial categories and three socio-economic status levels. Structural equation modeling was employed to test a hurdle regression model of dose-response to predict ITFV. A version of the Conflict Tactics Scale was used to measure physical violence victimization, witnessing parental IPV and young adult IPV perpetration and victimization. Results: Consistent with previous findings, past 12 months incidence rates severity and frequency of interpersonal violence were highly skewed. While rates of parental and young adult IPV were about 40%, an unusually high rate of physical child abuse (57%) was reported. The vast majority of a number of acts of violence, whether minor or severe, were in the 1-3 range in the past 12 months. Reported frequencies of more than 5 times in the past year were rare, with less than 10% of those reporting more than six acts of minor or severe physical violence. As expected, minor acts of violence were much more common than acts of severe violence. Overall, regression analyses were not significant for the dose-response model of ITFV. Conclusions and implications: The results of the dose-response model were not significant due to a lack of power in the final sample (N=338). Nonetheless, the value of the approach was confirmed for the future research given the bi-modal nature of the distributions which suggest that in the context of both child physical abuse and physical IPV, there are at least two classes when frequency of acts is considered. Taking frequency into account in predictive models may help to better understand the relationship of exposure to ITFV outcomes. Further testing using hurdle regression models is suggested.

Keywords: intergenerational transmission of family violence, physical child abuse, intimate partner violence, structural equation modeling

Procedia PDF Downloads 243

Authors: P. Kazymbet, G. Abildinova, K.Makhambetov, M. Bakhtin, D. Rybalkina, K. Zhumadilov

Abstract:

Conducted cytogenetic research in workers Stepnogorsk Mining-Chemical Combine (Akmola region) with the study of 26341 chromosomal metaphase. Using a regression analysis with program DataFit, version 5.0, dependence between exposure dose and the following cytogenetic exponents has been studied: frequency of aberrant cells, frequency of chromosomal aberrations, frequency of the amounts of dicentric chromosomes, and centric rings. Experimental data on calibration curves "dose-effect" enabled the development of a mathematical model, allowing on data of the frequency of aberrant cells, chromosome aberrations, the amounts of dicentric chromosomes and centric rings calculate the absorbed dose at the time of the study. In the dose range of 0.1 Gy to 5.0 Gy dependence cytogenetic parameters on the dose had the following equation: Y = 0,0067е^0,3307х (R2 = 0,8206) – for frequency of chromosomal aberrations; Y = 0,0057е^0,3161х (R2 = 0,8832) –for frequency of cells with chromosomal aberrations; Y =5 Е-0,5е^0,6383 (R2 = 0,6321) – or frequency of the amounts of dicentric chromosomes and centric rings on cells. On the basis of cytogenetic parameters and regression equations calculated absorbed dose in workers of uranium production at the time of the study did not exceed 0.3 Gy.

Keywords: Stepnogorsk, mathematical modeling, cytogenetic, dicentric chromosomes

Procedia PDF Downloads 478

Authors: Yibeltal T. Bayou, Yohana S. Mashalla, Gloria Thupayagale-Tshweneagae

Abstract:

Background: According to the World Health Organization, caesarean section delivery rates of more than 10-15% caesarean section deliveries in any specific geographic region in the world are not justifiable. The aim of the study was to describe the level and analyse determinants of caesarean section delivery in Addis Ababa. Methods: Data was collected in Addis Ababa using a structured questionnaire administered to 901 women aged 15-49 years through a stratified two-stage cluster sampling technique. Binary logistic regression model was employed to identify predictors of caesarean section delivery. Results: Among the 835 women who delivered their last birth at healthcare facilities, 19.2% of them gave birth by caesarean section. About 9.0% of the caesarean section births were due to mother’s request or service provider’s influence without any medical indication. The caesarean section delivery rate was much higher than the recommended rate particularly among the non-slum residents (27.2%); clients of private healthcare facilities (41.1%); currently married women (20.6%); women with secondary (22.2%) and tertiary (33.6%) level of education; and women belonging to the highest wealth quintile household (28.2%). The majority (65.8%) of the caesarean section clients were not informed about the consequences of caesarean section delivery by service providers. The logistic regression model shows that older age (30-49), secondary and above education, non-slum residence, high-risk pregnancy and receiving adequate antenatal care were significantly positively associated with caesarean section delivery. Conclusion: Despite the unreserved effort towards achieving MDG 5 through safe skilled delivery assistance among others, the high caesarean section rate beyond the recommend limit, and the finding that caesarean sections done without medical indications were also alarming. The government and city administration should take appropriate measures before the problems become setbacks in healthcare provision. Further investigations should focus on the effect of caesarean section delivery on maternal and child health outcomes in the study area.

Keywords: Addis Ababa, caesarean section, mode of delivery, slum residence

Procedia PDF Downloads 404

Authors: Sangeeta Palekar, Neeraj Rangwani, Akash Poddar, Jayu Kalambe

Abstract:

The bio-medical analysis is an indispensable procedure for identifying health-related diseases like diabetes. Monitoring the glucose level in our body regularly helps us identify hyperglycemia and hypoglycemia, which can cause severe medical problems like nerve damage or kidney diseases. This paper presents a method for predicting the glucose concentration in blood samples using image processing and machine learning algorithms. The glucose solution is prepared by the glucose oxidase (GOD) and peroxidase (POD) method. An experimental database is generated based on the colorimetric technique. The image of the glucose solution is captured by the raspberry pi camera and analyzed using image processing by extracting the RGB, HSV, LUX color space values. Regression algorithms like multiple linear regression, decision tree, RandomForest, and XGBoost were used to predict the unknown glucose concentration. The multiple linear regression algorithm predicts the results with 97% accuracy. The image processing and machine learning-based approach reduce the hardware complexities of existing platforms.

Keywords: artificial intelligence glucose detection, glucose oxidase, peroxidase, image processing, machine learning

Procedia PDF Downloads 204