Two-class Logistic Regression Research Articles

MicroRNA expression signature as a biomarker in the diagnosis of nodal T-cell lymphomas

BackgroundThe diagnosis of T-cell lymphomas is typically established through a multiparameter approach that combines clinical, morphologic, immunophenotypic, and genetic features, utilizing a variety of histopathologic and molecular techniques. However, accurate diagnosis of such lymphomas and distinguishing them from reactive lymph nodes remains challenging due to their low prevalence and heterogeneous features, hence limiting the confidence of pathologists. We investigated the use of microRNA (miRNA) expression signatures as an adjunctive tool in the diagnosis and classification of T-cell lymphomas that involve lymph nodes. This study seeks to distinguish reactive lymph nodes (RLN) from two types of frequently occurring nodal T-cell lymphomas: nodal T-follicular helper (TFH) cell lymphomas (nTFHL) and peripheral T-cell lymphomas, not otherwise specified (nPTCL).MethodsFrom the formalin-fixed paraffin-embedded (FFPE) samples from a cohort of 88 subjects, 246 miRNAs were quantified and analyzed by differential expression. Two-class logistic regression and random forest plot models were built to distinguish RLN from the nodal T-cell lymphomas. Gene set enrichment analysis was performed on the target genes of the miRNA to identify pathways and transcription factors that may be regulated by the differentially expressed miRNAs in each subtype.ResultsUsing logistic regression analysis, we identified miRNA signatures that can distinguish RLN from nodal T-cell lymphomas (AUC of 0.92 ± 0.05), from nTFHL (AUC of 0.94 ± 0.05) and from nPTCL (AUC of 0.94 ± 0.08). Random forest plot modelling was also capable of distinguishing between RLN and nodal T-cell lymphomas, but performed worse than logistic regression. However, the miRNA signatures are not able to discriminate between nTFHL and nPTCL, owing to large similarity in miRNA expression patterns. Bioinformatic analysis of the gene targets of unique miRNA expression revealed the enrichment of both known and potentially understudied signalling pathways and genes in such lymphomas.ConclusionThis study suggests that miRNA biomarkers may serve as a promising, cost-effective tool to aid the diagnosis of nodal T-cell lymphomas, which can be challenging. Bioinformatic analysis of differentially expressed miRNAs revealed both relevant or understudied signalling pathways that may contribute to the progression and development of each T-cell lymphoma entity. This may help us gain further insight into the biology of T-cell lymphomagenesis.

A Comparative Study Evaluated the Performance of Two-class Classification Algorithms in Machine Learning

Worldwide, heart attacks, also called myocardial infarctions, are a leading cause of death. Early detection and accurate prediction of heart attacks are crucial for effective medical intervention and patient care. In recent years, machine learning techniques have shown great promise in aiding the diagnosis and prediction of heart attacks. The Organization for World Health (WHO) reports that around 17 million individuals worldwide pass away from cardiovascular diseases (CVD), notably heart attacks and strokes, each year. In this study, 1026 patients, both men and women, are almost equally affected by CVDs. While heart attacks and strokes remain among the leading causes of mortality worldwide, the use of machine learning for predicting heart disease could potentially prevent premature deaths. A comparative study evaluated the performance of five well-known two-class classification algorithms: two-class boosted decision trees, two-class decision forests, two-class locally deep SVMs, two-class neural networks, and two-class logistic regression. Among these algorithms, the Two-Class Boosted Decision Tree method demonstrated outstanding prediction ability, achieving a 100% accuracy rating. Its exceptional recall and precision rates highlight its effectiveness in handling challenging classifications. To facilitate the development and deployment of machine learning models, Azure Machine Learning offers a range of tools and services. By leveraging Azure Machine Learning's capabilities, researchers and healthcare professionals can analyze large datasets containing patient information and medical records to identify patterns and risk factors associated with heart attacks.

Boutique Hotel Service Digitalization: A Business Owner Study

The COVID-19 pandemic has generated negative, economic impacts on the tourism and leisure sector in Thailand, especially small boutique hotels. These hotels have had to develop more efficient and innovative approaches to meet new normal expectations, for example, contactless service. Digital technologies, such as Machine Learning and Artificial Intelligence, can open new possibilities and opportunities for hotels to digitize their customers’ services. A review of the literature indicated that data important to the management of hotel products and services include Customer Segmentation, Customer Profiling, Menu Engineering, Productivity Indexing, Customer Associations, Forecasting, Energy Consumption, and Room Rates. These characteristics can be examined by machine learning. This study used a mixed qualitative and quantitative research method. The data were gathered by interviewing two boutique hotel owners in Bangkok and collecting the hotels’ data, including online travel booking agents and direct booking logs, for the period April 2016 – September 2021. The analysis was conducted using the booking data from the two hotels: 3946 records from Hotel A and 3948 from Hotel B. In this research, k-means clustering was used to segment hotel guests. Two-class logistic regression and a two-class boosted decision tree were used to predict the prospective customer, while linear regression and decision forest regression were used to forecast the market demand. The findings reveal a model of hotel business owners’ requirements to innovate new service solutions, such as the contactless software solution, that guests can employ for check-in, check-out, order services, and talk to the hotel through the mobile application. This would help hotel owners to manage costs, employees, and customers. The solution also means that hotel managers would no longer need to be involved in the manual implementation of revenue management tasks. This data analytics approach can effectively sift through the signals detected from market variables, discover patterns and anomalies, make predictions for guest arrivals, and calculate optimum prices in real-time, as the market changes.

Predicting the Onset of Diabetes with Machine Learning Methods.

The number of people suffering from diabetes in Taiwan has continued to rise in recent years. According to the statistics of the International Diabetes Federation, about 537 million people worldwide (10.5% of the global population) suffer from diabetes, and it is estimated that 643 million people will develop the condition (11.3% of the total population) by 2030. If this trend continues, the number will jump to 783 million (12.2%) by 2045. At present, the number of people with diabetes in Taiwan has reached 2.18 million, with an average of one in ten people suffering from the disease. In addition, according to the Bureau of National Health Insurance in Taiwan, the prevalence rate of diabetes among adults in Taiwan has reached 5% and is increasing each year. Diabetes can cause acute and chronic complications that can be fatal. Meanwhile, chronic complications can result in a variety of disabilities or organ decline. If holistic treatments and preventions are not provided to diabetic patients, it will lead to the consumption of more medical resources and a rapid decline in the quality of life of society as a whole. In this study, based on the outpatient examination data of a Taipei Municipal medical center, 15,000 women aged between 20 and 80 were selected as the subjects. These women were patients who had gone to the medical center during 2018-2020 and 2021-2022 with or without the diagnosis of diabetes. This study investigated eight different characteristics of the subjects, including the number of pregnancies, plasma glucose level, diastolic blood pressure, sebum thickness, insulin level, body mass index, diabetes pedigree function, and age. After sorting out the complete data of the patients, this study used Microsoft Machine Learning Studio to train the models of various kinds of neural networks, and the prediction results were used to compare the predictive ability of the various parameters for diabetes. Finally, this study found that after comparing the models using two-class logistic regression as well as the two-class neural network, two-class decision jungle, or two-class boosted decision tree for prediction, the best model was the two-class boosted decision tree, as its area under the curve could reach a score of 0.991, which was better than other models.

IoT and Fog-Computing-Based Predictive Maintenance Model for Effective Asset Management in Industry 4.0 Using Machine Learning

The assets in Industry 4.0 are categorized into physical, virtual, and human. The innovation and popularization of ubiquitous computing enhance the usage of smart devices: RFID tags, QR codes, LoRa tags, etc., for asset identification and tracking. The generated data from the Industrial Internet of Things (IIoT) ease information visibility and process automation in Industry 4.0. Virtual assets include the data produced from IIoT. One of the applications of the industrial big data is to predict the failure of the manufacturing equipment. Predictive maintenance enables the business owner to decide, such as repairing or replacing the component before an actual failure that affects the whole production line. Therefore, Industry 4.0 requires an effective asset management to optimize the task distributions and predictive maintenance model. This article presents the genetic algorithm (GA)-based resource management integrating with machine learning for predictive maintenance in fog computing. The time, cost, and energy performance of GA along with MinMin, MaxMin, FCFS, and RoundRobin are simulated in the FogWorkflowsim. The predictive maintenance model is built in two-class logistic regression using real-time data sets. The results demonstrate that the proposed technique outperforms MinMin, MaxMin, FCFS, RoundRobin in execution time, cost, and energy usage. The execution time is 0.48% faster, 5.43% lower cost and energy usage is 28.10% lower in comparison with second-best results. The training and testing accuracy of the prediction model is 95.1% and 94.5%, respectively.

Current situation and influencing factors of the irritability of the elderly in Mianyang City in 2019

To understand the current situation of irritability in the elderly in Mianyang City, explore the influencing factors of irritability. In 2019, 823 elderly people over 60 years old in Mianyang City were selected by multi-stage random sampling, including 351(42.65%) males and 472(57.35%) females, with a median age of 67 years and a interquartile interval of 8 years. The information was collected by self rating scale of irritability, depression and anxiety and questionnaire of influencing factors of negative emotion. Questionnaire surveys were used to investigate their irritability, health status, health-related behaviors, support for the elderly, characteristics of the elderly, and mental health knowledge. Describe the detection rate of irritability, and use two-class logistic regression model to analyze the influencing factors of their depressive symptoms. A total of 823 elderly people in Mianyang City were investigated, 178 were found to be irritable, and the prevalence of irritability was 21.63%. The result of Logistic regression model showed that: female(OR= 1.800, 95% CI 1.223-2.627), ill within two weeks(OR = 1.611, 95% CI 1.101-2.358), hospitalized within one year(OR = 2.045, 95% CI 1.327-3.048), no spouse(OR = 0.349, 95% CI 0.209-0.582), good self-rated health(OR = 0.389, 95% CI 0.243-0.625), emotional support for children(OR = 0.549, 95% CI 0.341-0.882), the old-age care model was self care(OR = 0.649, 95% CI 0.453-0.931), they were related to irritating emotions. A certain proportion of the elderly in Mianyang City have irritability. Gender, marital status, health status, pension support and characteristics are the influencing factors of irritability.

Risk factors of mortality in peritoneal dialysis patients

Objective To explore the risk factors affecting the death of peritoneal dialysis patients. Methods The clinical data of 275 patients with peritoneal dialysis who place a peritoneal catheter in the Department of Nephrology, First Affiliated Hospital of Guangxi Medical University and had regular peritoneal dialysis for more than 3 months from January 1, 2005 to December 31, 2015 were retrospectively analyzed.Sixty patients died in the observation group, and 215 patients who continued regular peritoneal dialysis were in the control group.Univariate analysis and two-class logistic regression analysis were used to analyze risk factors for death. Results The composition ratio of the primary disease to diabetic nephropathy in the death group and the control group was 15.0% (9/60) and 5.6% (12/215), respectively, the average age of patients entering peritoneal dialysis was (50.6±14.3) years old and (45.7±13.2) years old, respectively(t=-2.518), glomerular filtration rates were 6.0(4.5, 9.4) and 5.1(4.2, 6.6), respectively, blood potassium is (4.2±0.7) mmol/L and (4.5±0.7) mmol/L, respectively, serum creatinine was 721.0 (585.0, 891.3) μmol/L and 847 (723.3, 1 033.3) μmol/L, respectively, there were significant differences between the two groups (t=2.14, all P<0.05). Logistic regression analysis indicated that the primary disease was diabetic nephropathy, and the age of admission to peritoneal dialysis, glomerular filtration rate, blood uric acid are risk factors for death in peritoneal dialysis patients. Conclusion Diabetic nephropathy, age, glomerular filtration rate, and blood uric acid level are independent risk factors for death in peritoneal dialysis patients. Key words: Peritoneal dialysis; Glomerular filtration rate; Risk factors

Medical staff's exposure to solid particles and health impact survey

Objective: To investigate the current situation of exposure of solid particles in medical staff and the occurrence of related diseases, and to understand the distribution characteristics of occupational exposure and the influencing factors. Methods: In August 2016, the stratified cluster sampling method was adopted to select three tertiary general hospitals with different regions, different administrative levels and departments in Shandong Province. A face-to-face questionnaire survey was conducted on the basic situation of the staff of the hospital, the status of exposure to occupational hazards and the health effects. Two-class logistic regression analysis was used to analyze the relationship between medical staff's solid particle contact and occupational health effects. In August 2016, the stratified cluster sampling method was adopted to select three tertiary general hospitals with different regions, different administrative levels and departments in Shandong Province, which were in contact with the basic conditions and occupational hazards of the staff of the hospital. Current situation and health impacts were conducted face-to-face questionnaires. Two-class logistic regression analysis was used to analyze the relationship between medical staff's solid particle contact and occupational health effects. Results: The overall contact rate of solid particles of medical staff was 41.5% (1843/4 446) , and the solid particles with the highest contact rate were 28.2% (1254/4 446) of talc dust. There was a statistically significant difference in the contact rate of solid particles between medical staff of different occupations, current working age, education and hospital type (χ(2)=76.30, 41.26, 26.16, 188.97, P<0.05) . There were significant differences in the conditions and types of solid particles in different departments (χ(2)=740.23, P<0.05) . Compared with the control group, the incidence of allergic asthma, chronic tracheobronchitis and dermatitis was higher in the contact group medical staff (P<0.05) . Contact with zirconia dust, talc dust, cotton dust is a risk factor for dermatitis in medical staff (OR=2.09, 1.45, 1.34) ; exposure to zirconia dust and cotton dust is a risk factor for chronic tracheobronchitis in medical staff (OR=2.38, 1.66) ; exposure to zirconia dust, gypsum dust is a risk factor for allergic asthma in medical staff (OR=3.90, 1.52) . Conclusion: The contact rate of solid particles in medical staff is high, and there are many types of contact. Frequent contacted with solid particles is a dangerous cause of related diseases. The working environment of medical staff should be further monitored and be strengthen the targeted protective measures and regulatory systems for the exposure of medical personnel to solid particles.

The preliminary study on the combined application model of serum miRNAs in the diagnosis of pneumoconiosis

Objective: To explore the role of serum microRNA (miRNAs) levels in the detection of pneumoconiosis, and to establish a combined application model of multiple serums miRNAs for pneumoconiosis diagnosis. Methods: 152 cases were selected in the case group and the control group respectively. The TaqMan Low Density Array method was used to screen out the candidate miRNAs for early screening of pneumoconiosis, and RT-qPCR was used to verify. According to the area under the curve (AUC) , the sensitivity and specificity of the candidate indicators were investigated. The logistic regression model was established by the two-class logistic regression model. Results: The expression of 7 candidate miRNAs in the serum of pneumoconiosis patients was significantly different (P<0.05) . The receiver operating curve (ROC) of the above 7 miRNAs was analyzed, miRNA-21, miRNA-200c, miRNA-16, miRNA-206, miRNA-155, miRNA-29a had statistical significance, and their ROC-AUC is 0.629~0.932. Logistic regression model was: logitP=13.769+0.536×miRNA-21-0.878×miRNA-200C-0.012×miRNA-16-0.111×miRNA-206+0.117×miRNA-155-1.192×miRNA-29a. Conclusion: Multiple serum miRNAs combined application models may be used for the diagnosis of pneumoconiosis patients.

Regularization Paths for Generalized Linear Models via Coordinate Descent

We develop fast algorithms for estimation of generalized linear models with convex penalties. The models include linear regression, two-class logistic regression, and multinomial regression problems while the penalties include ℓ(1) (the lasso), ℓ(2) (ridge regression) and mixtures of the two (the elastic net). The algorithms use cyclical coordinate descent, computed along a regularization path. The methods can handle large problems and can also deal efficiently with sparse features. In comparative timings we find that the new algorithms are considerably faster than competing methods.