A machine learning-based approach to determine infection status in recipients of BBV152 whole virion inactivated SARS-CoV-2 vaccine for serological surveys

Data science has been an invaluable part of the COVID-19 pandemic response with multiple applications, ranging from tracking viral evolution to understanding the effectiveness of interventions. Asymptomatic breakthrough infections have been a major problem during the ongoing surge of Delta variant globally. Serological discrimination of vaccine response from infection has so far been limited to Spike protein vaccines used in the higher-income regions. Here, we show for the first time how statistical and machine learning (ML) approaches can discriminate SARS-CoV-2 infection from immune response to an inactivated whole virion vaccine (BBV152, Covaxin, India), thereby permitting real-world vaccine effectiveness assessments from cohort-based serosurveys in Asia and Africa where such vaccines are commonly used. Briefly, we accessed serial data on Anti-S and Anti-NC antibody concentration values, along with age, sex, number of doses, and number of days since the last vaccine dose for 1823 Covaxin recipients. An ensemble ML model, incorporating a consensus clustering approach alongside the support vector machine (SVM) model, was built on 1063 samples where reliable qualifying data existed, and then applied to the entire dataset. Of 1448 self-reported negative subjects, 724 were classified as infected. Since the vaccine contains wild-type virus and the antibodies induced will neutralize wild type much better than Delta variant, we determined the relative ability of a random subset of such samples to neutralize Delta versus wild type strain. In 100 of 156 samples, where ML prediction differed from self-reported uninfected status, Delta variant, was neutralized more effectively than the wild type, which cannot happen without infection. The fraction rose to 71.8% (28 of 39) in subjects predicted to be infected during the surge, which is concordant with the percentage of sequences classified as Delta (75.6%-80.2%) over the same period.

Download Full-text

Detection of novel HIV-1 drug resistance mutations by support vector analysis of deep sequence data and experimental validation

10.1101/804781 ◽

2019 ◽

Author(s):

Mariano Avino ◽

Emmanuel Ndashimye ◽

Daniel J. Lizotte ◽

Abayomi S. Olabode ◽

Richard M. Gibson ◽

...

Keyword(s):

Drug Resistance ◽

Drug Treatment ◽

Treatment Outcomes ◽

Sequence Data ◽

Wild Type Virus ◽

Support Vector ◽

Wild Type ◽

Subtype B ◽

The World ◽

Hiv 1

AbstractThe global HIV-1 pandemic comprises many genetically divergent subtypes. Most of our understanding of drug resistance in HIV-1 derives from subtype B, which predominates in North America and western Europe. However, about 90% of the pandemic represents non-subtype B infections. Here, we use deep sequencing to analyze HIV-1 from infected individuals in Uganda who were either treatment-naïve or who experienced virologic failure on ART without the expected patterns of drug resistance. Our objective was to detect potentially novel associations between mutations in HIV-1 integrase and treatment outcomes in Uganda, where most infections are subtypes A or D. We retrieved a total of 380 archived plasma samples from patients at the Joint Clinical Research Centre (Kampala), of which 328 were integrase inhibitor-naïve and 52 were raltegravir (RAL)-based treatment failures. Next, we developed a bioinformatic pipeline for alignment and variant calling of the deep sequence data obtained from these samples from a MiSeq platform (Illumina). To detect associations between within-patient polymorphisms and treatment outcomes, we used a support vector machine (SVM) for feature selection with multiple imputation to account for partial reads and low quality base calls. Candidate point mutations of interest were experimentally introduced into the HIV-1 subtype B NL4-3 backbone to determine susceptibility to RAL in U87.CD4.CXCR4 cells. Finally, we carried out replication capacity experiments with wild-type and mutant viruses in TZM-bl cells in the presence and absence of RAL. Our analyses not only identified the known major mutation N155H and accessory mutations G163R and V151I, but also novel mutations I203M and I208L as most highly associated with RAL failure. The I203M and I208L mutations resulted in significantly decreased susceptibility to RAL (44.0-fold and 54.9-fold, respectively) compared to wild-type virus (EC50=0.32 nM), and may represent novel pathways of HIV-1 resistance to modern treatments.Author summaryThere are many different types of HIV-1 around the world. Most of the research on how HIV-1 can become resistant to drug treatment has focused on the type (B) that is the most common in high-income countries. However, about 90% of infections around the world are caused by a type other than B. We used next-generation sequencing to analyze samples of HIV-1 from patients in Uganda (mostly infected by types A and D) for whom drug treatment failed to work, and whose infections did not fit the classic pattern of adaptation based on B. Next, we used machine learning to detect mutations in these virus populations that could explain the treatment outcomes. Finally, we experimentally added two candidate mutations identified by our analysis to a laboratory strain of HIV-1 and confirmed that they conferred drug resistance to the virus. Our study reveals new pathways that other types of HIV-1 may use to evolve resistance to drugs that make up the current recommended treatment for newly diagnosed individuals.

Download Full-text

Diabetes Prediction Using Machine Learning

International Journal of Scientific Research in Computer Science Engineering and Information Technology ◽

10.32628/cseit206463 ◽

2020 ◽

pp. 294-305

Author(s):

KM Jyoti Rani

Keyword(s):

Machine Learning ◽

Blood Glucose ◽

Data Science ◽

International Diabetes Federation ◽

Machine Learning Techniques ◽

Support Vector ◽

High Blood Glucose ◽

Diabetes Prediction ◽

The One ◽

Prediction Of Diabetes

Diabetes is a chronic disease with the potential to cause a worldwide health care crisis. According to International Diabetes Federation 382 million people are living with diabetes across the whole world. By 2035, this will be doubled as 592 million. Diabetes is a disease caused due to the increase level of blood glucose. This high blood glucose produces the symptoms of frequent urination, increased thirst, and increased hunger. Diabetes is a one of the leading cause of blindness, kidney failure, amputations, heart failure and stroke. When we eat, our body turns food into sugars, or glucose. At that point, our pancreas is supposed to release insulin. Insulin serves as a key to open our cells, to allow the glucose to enter and allow us to use the glucose for energy. But with diabetes, this system does not work. Type 1 and type 2 diabetes are the most common forms of the disease, but there are also other kinds, such as gestational diabetes, which occurs during pregnancy, as well as other forms. Machine learning is an emerging scientific field in data science dealing with the ways in which machines learn from experience. The aim of this project is to develop a system which can perform early prediction of diabetes for a patient with a higher accuracy by combining the results of different machine learning techniques. The algorithms like K nearest neighbour, Logistic Regression, Random forest, Support vector machine and Decision tree are used. The accuracy of the model using each of the algorithms is calculated. Then the one with a good accuracy is taken as the model for predicting the diabetes.

Download Full-text

Classification of different skarn deposits based on the compositional variability of associated grandite garnets: a data science and Machine Learning approach

10.5194/egusphere-egu21-10537 ◽

2021 ◽

Author(s):

Urmi Ghosh ◽

Tuhin Chakraborty

Keyword(s):

Machine Learning ◽

Trace Element ◽

Data Science ◽

Training Dataset ◽

Support Vector ◽

Learning Approach ◽

Machine Learning Approach ◽

Skarn Deposits

<p>Rapid technological improvements made in in-situ analysis techniques, including LA-ICPMS, have transformed the field of analytical geochemistry. This has a far-reaching impact for different petrogenetic and ore-genetic studies where minute major and trace element compositional changes between different mineral zones within a single crystal can now be demarcated. Minerals such as garnet although robust are quite sensitive to the changing P-T and fluid conditions during their formation. These minerals have become powerful tools to characterize mineralization types. Previously, Meinert (1992) has used in-situ major element EPMA analysis results to classify different skarn deposit based on the end-member composition of hydrothermal garnets. Alternatively, Tian et al. (2019) used the garnet trace element composition for the similar purpose. However, these discrimination plots/ classification schemes show major overlap in different skarn deposits, such as Fe, Cu, Zn, and Au. The present study is an attempt to use machine learning approach on available garnet data to found a more potent classification scheme for skarn deposits, thus reaffirming garnet as a faithful indicator for hydrothermal ore deposits. We have meticulously collected major and trace element data of Ca-rich garnets, associated with different skarn deposits worldwide from 40 publications. This collected data is then used to train a model for fingerprinting the skarn deposits. Stratified random sampling method has been used on the dataset with 80% of the samples as test set and the rest 20 % as training dataset. We have used K-nearest neighbour (KNN), Support Vector Machine (SVM) and Random Forest algorithms on the data by using Python as a platform. These ML classification algorithm performs better than the earlier existing models available for classification of ore types based on garnet composition in skarn system. Factor importance is calculated that shows which elements play a pivotal role in classification of the ore type. Our results depict that multiple garnet forming elements taken together can reliably be used to discriminate between different ore formation settings.</p>

Download Full-text

Earthquake Prediction using Machine Learning Algorithm

International Journal of Recent Technology and Engineering - 2 ◽

10.35940/ijrte.e9110.018620 ◽

2020 ◽

Vol 8 (6) ◽

pp. 4684-4688

Keyword(s):

Machine Learning ◽

Structural Damage ◽

Data Science ◽

Learning Algorithm ◽

Economic Loss ◽

Machine Learning Algorithms ◽

Training Data ◽

Support Vector ◽

Science Data ◽

Data Set

Per the statistics received from BBC, data varies for every earthquake occurred till date. Approximately, up to thousands are dead, about 50,000 are injured, around 1-3 Million are dislocated, while a significant amount go missing and homeless. Almost 100% structural damage is experienced. It also affects the economic loss, varying from 10 to 16 million dollars. A magnitude corresponding to 5 and above is classified as deadliest. The most life-threatening earthquake occurred till date took place in Indonesia where about 3 million were dead, 1-2 million were injured and the structural damage accounted to 100%. Hence, the consequences of earthquake are devastating and are not limited to loss and damage of living as well as nonliving, but it also causes significant amount of change-from surrounding and lifestyle to economic. Every such parameter desiderates into forecasting earthquake. A couple of minutes’ notice and individuals can act to shield themselves from damage and demise; can decrease harm and monetary misfortunes, and property, characteristic assets can be secured. In current scenario, an accurate forecaster is designed and developed, a system that will forecast the catastrophe. It focuses on detecting early signs of earthquake by using machine learning algorithms. System is entitled to basic steps of developing learning systems along with life cycle of data science. Data-sets for Indian sub-continental along with rest of the World are collected from government sources. Pre-processing of data is followed by construction of stacking model that combines Random Forest and Support Vector Machine Algorithms. Algorithms develop this mathematical model reliant on “training data-set”. Model looks for pattern that leads to catastrophe and adapt to it in its building, so as to settle on choices and forecasts without being expressly customized to play out the task. After forecast, we broadcast the message to government officials and across various platforms. The focus of information to obtain is keenly represented by the 3 factors – Time, Locality and Magnitude.

Download Full-text

PREDICCIÓN DE FUGA DE CLIENTES EN UNA EMPRESA DE DISTRIBUCIÓN DE GAS NATURAL MEDIANTE EL USO DE MINERÍA DE DATOS

Universidad Ciencia y Tecnología ◽

10.47460/uct.v24i106.399 ◽

2020 ◽

Vol 24 (106) ◽

pp. 79-87

Author(s):

Fredy Humberto Troncoso Espinosa ◽

Javiera Valentina Ruiz Tapia

Keyword(s):

Machine Learning ◽

Data Mining ◽

Feature Selection ◽

Knowledge Discovery ◽

Data Science ◽

Machine Learning Techniques ◽

Support Vector ◽

Churn Prediction ◽

Customer Churn ◽

Using Data

La fuga de clientes es un problema relevante al que enfrentan las empresas de servicios y que les puede generar pérdidas económicas significativas. Identificar los elementos que llevan a un cliente a dejar de consumir un servicio es una tarea compleja, sin embargo, mediante su comportamiento es posible estimar una probabilidad de fuga asociada a cada uno de ellos. Esta investigación aplica minería de datos para la predicción de la fuga de clientes en una empresa de distribución de gas natural, mediante dos técnicas de machine learning: redes neuronales y support vector machine. Los resultados muestran que mediante la aplicación de estas técnicas es posible identificar los clientes con mayor probabilidad de fuga para tomar sobre estas acciones de retenciónoportunas y focalizadas, minimizando los costos asociados al error en la identificación de estos clientes. Palabras Clave: fuga de clientes, minería de datos, machine learning, distribución de gas natural. Referencias [1]J. Miranda, P. Rey y R. Weber, «Predicción de Fugas de Clientes para una Institución Financiera Mediante Support Vector Machines,» Revista Ingeniería de Sistemas Volumen XIX, pp. 49-68, 2005. [2]P. A. Pérez V., «Modelo de predicción de fuga de clientes de telefonía movil post pago,» Universidad de Chile, Santiago, Chile, 2014. [3]Gas Sur S.A., «https://www.gassur.cl/Quienes-Somos/,» [En línea]. [4]J. Xiao, X. Jiang, C. He y G. Teng, «Churn prediction in customer relationship management via GMDH-based multiple classifiers ensemble,» IEEE IntelligentSystems, vol. 31, nº 2, pp. 37-44, 2016. [5]A. M. Almana, M. S. Aksoy y R. Alzahrani, «A survey on data mining techniques in customer churn analysis for telecom industry,» International Journal of Engineering Research and Applications, vol. 4, nº 5, pp. 165-171, 2014. [6]A. Jelvez, M. Moreno, V. Ovalle, C. Torres y F. Troncoso, «Modelo predictivo de fuga de clientes utilizando mineríaa de datos para una empresa de telecomunicaciones en chile,» Universidad, Ciencia y Tecnología, vol. 18, nº 72, pp. 100-109, 2014. [7]D. Anil Kumar y V. Ravi, «Predicting credit card customer churn in banks using data mining,» International Journal of Data Analysis Techniques and Strategies, vol. 1, nº 1, pp. 4-28, 2008. [8]E. Aydoğan, C. Gencer y S. Akbulut, «Churn analysis and customer segmentation of a cosmetics brand using data mining techniques,» Journal of Engineeringand Natural Sciences, vol. 26, nº 1, 2008. [9]G. Dror, D. Pelleg, O. Rokhlenko y I. Szpektor, «Churn prediction in new users of Yahoo! answers,» de Proceedings of the 21st International Conference onWorld Wide Web, 2012. [10]T. Vafeiadis, K. Diamantaras, G. Sarigiannidis y K. Chatzisavvas, «A comparison of machine learning techniques for customer churn prediction,» SimulationModelling Practice and Theory, vol. 55, pp. 1-9, 2015. [11]Y. Xie, X. Li, E. Ngai y W. Ying, «Customer churn prediction using improved balanced random forests,» Expert Systems with Applications, vol. 36, nº 3, pp.5445-5449, 2009. [12]U. Fayyad, G. Piatetsky-Shapiro y P. Smyth, «Knowledge Discovery and Data Mining: Towards a Unifying Framework,» de KDD-96 Proceedings, 1996. [13]R. Brachman y T. Anand, «The process of knowledge discovery in databases,» de Advances in knowledge discovery and data mining, 1996. [14]K. Lakshminarayan, S. Harp, R. Goldman y T. Samad, «Imputation of Missing Data Using Machine Learning Techniques,» de KDD, 1996. [15]B. Nguyen , J. L. Rivero y C. Morell, «Aprendizaje supervisado de funciones de distancia: estado del arte,» Revista Cubana de Ciencias Informáticas, vol. 9, nº 2, pp. 14-28, 2015. [16]I. Monedero, F. Biscarri, J. Guerrero, M. Peña, M. Roldán y C. León, «Detection of water meter under-registration using statistical algorithms,» Journal of Water Resources Planning and Management, vol. 142, nº 1, p. 04015036, 2016. [17]I. Guyon y A. Elisseeff, «An introduction to variable and feature selection,» Journal of machine learning research, vol. 3, nº Mar, pp. 1157-1182, 2003. [18]K. Polat y S. Güneş, «A new feature selection method on classification of medical datasets: Kernel F-score feature selection,» Expert Systems with Applications, vol. 36, nº 7, pp. 10367-10373, 2009. [19]D. J. Matich, «Redes Neuronales. Conceptos Básicos y Aplicaciones,» de Cátedra: Informática Aplicada ala Ingeniería de Procesos- Orientación I, 2001. [20]E. Acevedo M., A. Serna A. y E. Serna M., «Principios y Características de las Redes Neuronales Artificiales, » de Desarrollo e Innovación en Ingeniería, Medellín, Editorial Instituto Antioqueño de Investigación, 2017, pp. Capítulo 10, 173-182. [21]M. Hofmann y R. Klinkenberg, RapidMiner: Data mining use cases and business analytics applications, CRC Press, 2016. [22]R. Pupale, «Towards Data Science,» 2018. [En línea]. Disponible: https://towardsdatascience.com/https-medium-com-pupalerushikesh-svm-f4b42800e989. [23]F. H. Troncoso Espinosa, «Prediction of recidivismin thefts and burglaries using machine learning,» Indian Journal of Science and Technology, vol. 13, nº 6, pp. 696-711, 2020. [24]L. Tashman, «Out-of-sample tests of forecasting accuracy: an analysis and review,» International journal of forecasting, vol. 16, nº 4, pp. 437-450, 2000. [25]S. Varma y R. Simon, «Bias in error estimation when using cross-validation for model selection,» BMC bioinformatics, vol. 7, nº 1, p. 91, 2006. [26]N. V. Chawla, K. W. Bowyer, L. O. Hall y W. Kegelmeyer, «SMOTE: Synthetic Minority Over-sampling Technique,» Journal of Artificial Inteligence Research16, pp. 321-357, 2002. [27]M. Sokolova y G. Lapalme, «A systematic analysis of performance measures for classification tasks,» Information processing & management, vol. 45, nº 4, pp. 427-437, 2009. [28]S. Narkhede, «Understanding AUC-ROC Curve,» Towards Data Science, vol. 26, 2018. [29]R. Westermann y W. Hager, «Error Probabilities in Educational and Psychological Research,» Journal of Educational Statistics, Vol 11, No 2, pp. 117-146, 1986.

Download Full-text

A review of machine learning applications in wildfire science and management

Environmental Reviews ◽

10.1139/er-2020-0019 ◽

2020 ◽

Vol 28 (4) ◽

pp. 478-505 ◽

Cited By ~ 5

Author(s):

Piyush Jain ◽

Sean C.P. Coogan ◽

Sriram Ganapathi Subramanian ◽

Mark Crowley ◽

Steve Taylor ◽

...

Keyword(s):

Machine Learning ◽

Neural Networks ◽

Deep Learning ◽

Data Science ◽

Multiple Scales ◽

Fire Behavior ◽

Fire Detection ◽

Active Role ◽

Fire Effects ◽

Support Vector

Artificial intelligence has been applied in wildfire science and management since the 1990s, with early applications including neural networks and expert systems. Since then, the field has rapidly progressed congruently with the wide adoption of machine learning (ML) methods in the environmental sciences. Here, we present a scoping review of ML applications in wildfire science and management. Our overall objective is to improve awareness of ML methods among wildfire researchers and managers, as well as illustrate the diverse and challenging range of problems in wildfire science available to ML data scientists. To that end, we first present an overview of popular ML approaches used in wildfire science to date and then review the use of ML in wildfire science as broadly categorized into six problem domains, including (i) fuels characterization, fire detection, and mapping; (ii) fire weather and climate change; (iii) fire occurrence, susceptibility, and risk; (iv) fire behavior prediction; (v) fire effects; and (vi) fire management. Furthermore, we discuss the advantages and limitations of various ML approaches relating to data size, computational requirements, generalizability, and interpretability, as well as identify opportunities for future advances in the science and management of wildfires within a data science context. In total, to the end of 2019, we identified 300 relevant publications in which the most frequently used ML methods across problem domains included random forests, MaxEnt, artificial neural networks, decision trees, support vector machines, and genetic algorithms. As such, there exists opportunities to apply more current ML methods — including deep learning and agent-based learning — in the wildfire sciences, especially in instances involving very large multivariate datasets. We must recognize, however, that despite the ability of ML models to learn on their own, expertise in wildfire science is necessary to ensure realistic modelling of fire processes across multiple scales, while the complexity of some ML methods such as deep learning requires a dedicated and sophisticated knowledge of their application. Finally, we stress that the wildfire research and management communities play an active role in providing relevant, high-quality, and freely available wildfire data for use by practitioners of ML methods.

Download Full-text

OS1.7 Plasmatic differential scanning fluorimetry profiles discriminate glioma patients from healthy controls

Neuro-Oncology ◽

10.1093/neuonc/noz126.020 ◽

2019 ◽

Vol 21 (Supplement_3) ◽

pp. iii7-iii7

Author(s):

E Tabouret ◽

P Tsvetkov ◽

R Eyraud ◽

T Peel ◽

S Malesinski ◽

...

Keyword(s):

Machine Learning ◽

Protein Denaturation ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Support Vector ◽

Healthy Controls ◽

Wild Type ◽

Protein Fluorescence ◽

Adaptive Boosting ◽

Differential Scanning Fluorimetry

Abstract BACKGROUND differential scanning fluorimetry (DSF) has been recently proposed to be used to perform high throughput biofluids profiling by following protein denaturation. Our objective was to discriminate patients with glioma from healthy controls using plasmatic DSF profiles. MATERIAL AND METHODS We included 78 glioma patients and 44 healthy controls. Plasmas were collected using EDTA tubes and analyzed in duplicate using nanoDSF Prometheus NT.Plex instrument (Nanotemper). The following DSF data were analyzed: protein fluorescence at 330 and 350 nm, first derivation of the ratio of fluorescence at 330 and 350 and the absorbance at 350 nm. Then we ran several machine learning algorithms to differentiate gliomas from healthy controls: Logistic Regression (LR), Support Vector Machine (SVM), Neural Networks (NN), Random Forest (RF) and Adaptive Boosting (AdaBoost). All these methods have been tested using a leave-one-out approach where each datum is used once as test while the other are used to train the automatic classifiers. RESULTS We included 78 patients with a median age of 57.8 years (range, 21.8 - 89.9). Thirteen patients (17%) presented with a 1p/19q codeleted IDH mutated oligodendroglioma, 12 patients (15%) with an IDH mutated astrocytoma and 53 patients (68%) with an IDH wild-type astrocytoma, including 26 patients with a recurrent grade IV IDH wild-type astrocytoma. DSF data were independent from classical prognostic factors or patient characteristics: age, Karnofsky Performans Status, IDH mutation status, 1p19q codeletion, grade, initial or recurrent setting, steroid doses, patient size and weight, tumor size and location. The different datasets of the DFS output were tested independently and in combination as input of the machine learning algorithms. Results were obtained using the tuned best parameters on 157 data. The best obtained accuracy was 95.54% with 2% of fake positives and 5% of fake negatives (algorithm: SVM). Others achieved ≥ 90% of correct classification: LR accuracy was 89.17%, NN accuracy was 92.99%, RF accuracy was 91.72% and Adaboost accuracy was 92.36%. CONCLUSION DSF profiles analyzed by machine learning algorithms could allow us to identify glioma patients from healthy controls with an accuracy of more than 95%. These results suggest that DSF of biofluid could be a useful and non-invasive tool to monitor glioma patients. Further investigations, including longitudinal profile analyses are ongoing.

Download Full-text

What can machines learn about heart failure? A systematic literature review

International Journal of Data Science and Analytics ◽

10.1007/s41060-021-00300-1 ◽

2021 ◽

Author(s):

A. Jasinska-Piadlo ◽

R. Bond ◽

P. Biglarbeigi ◽

R. Brisk ◽

P. Campbell ◽

...

Keyword(s):

Machine Learning ◽

Heart Failure ◽

Clinical Practice ◽

Literature Review ◽

Systematic Literature Review ◽

Data Science ◽

Prediction Models ◽

Critical Evaluation ◽

Support Vector ◽

Clustering Methods

AbstractThis paper presents a systematic literature review with respect to application of data science and machine learning (ML) to heart failure (HF) datasets with the intention of generating both a synthesis of relevant findings and a critical evaluation of approaches, applicability and accuracy in order to inform future work within this field. This paper has a particular intention to consider ways in which the low uptake of ML techniques within clinical practice could be resolved. Literature searches were performed on Scopus (2014-2021), ProQuest and Ovid MEDLINE databases (2014-2021). Search terms included ‘heart failure’ or ‘cardiomyopathy’ and ‘machine learning’, ‘data analytics’, ‘data mining’ or ‘data science’. 81 out of 1688 articles were included in the review. The majority of studies were retrospective cohort studies. The median size of the patient cohort across all studies was 1944 (min 46, max 93260). The largest patient samples were used in readmission prediction models with the median sample size of 5676 (min. 380, max. 93260). Machine learning methods focused on common HF problems: detection of HF from available dataset, prediction of hospital readmission following index hospitalization, mortality prediction, classification and clustering of HF cohorts into subgroups with distinctive features and response to HF treatment. The most common ML methods used were logistic regression, decision trees, random forest and support vector machines. Information on validation of models was scarce. Based on the authors’ affiliations, there was a median 3:1 ratio between IT specialists and clinicians. Over half of studies were co-authored by a collaboration of medical and IT specialists. Approximately 25% of papers were authored solely by IT specialists who did not seek clinical input in data interpretation. The application of ML to datasets, in particular clustering methods, enabled the development of classification models assisting in testing the outcomes of patients with HF. There is, however, a tendency to over-claim the potential usefulness of ML models for clinical practice. The next body of work that is required for this research discipline is the design of randomised controlled trials (RCTs) with the use of ML in an intervention arm in order to prospectively validate these algorithms for real-world clinical utility.

Download Full-text

Using Machine Learning to Predict Heart Disease

WSEAS TRANSACTIONS ON BIOLOGY AND BIOMEDICINE ◽

10.37394/23208.2022.19.1 ◽

2022 ◽

Vol 19 ◽

pp. 1-9

Author(s):

Nikhil Bora ◽

Sreedevi Gutta ◽

Ahmad Hadaegh

Keyword(s):

Machine Learning ◽

Support Vector Machine ◽

Heart Disease ◽

Random Forest ◽

Data Science ◽

Learning Algorithm ◽

Machine Learning Algorithms ◽

Machine Learning Techniques ◽

Support Vector ◽

K Nearest Neighbor

Heart Disease has become one of the most leading cause of the death on the planet and it has become most life-threatening disease. The early prediction of the heart disease will help in reducing death rate. Predicting Heart Disease has become one of the most difficult challenges in the medical sector in recent years. As per recent statistics, about one person dies from heart disease every minute. In the realm of healthcare, a massive amount of data was discovered for which the data-science is critical for analyzing this massive amount of data. This paper proposes heart disease prediction using different machine-learning algorithms like logistic regression, naïve bayes, support vector machine, k nearest neighbor (KNN), random forest, extreme gradient boost, etc. These machine learning algorithm techniques we used to predict likelihood of person getting heart disease on the basis of features (such as cholesterol, blood pressure, age, sex, etc. which were extracted from the datasets. In our research we used two separate datasets. The first heart disease dataset we used was collected from very famous UCI machine learning repository which has 303 record instances with 14 different attributes (13 features and one target) and the second dataset that we used was collected from Kaggle website which contained 1190 patient’s record instances with 11 features and one target. This dataset is a combination of 5 popular datasets for heart disease. This study compares the accuracy of various machine learning techniques. In our research, for the first dataset we got the highest accuracy of 92% by Support Vector Machine (SVM). And for the second dataset, Random Forest gave us the highest accuracy of 94.12%. Then, we combined both the datasets which we used in our research for which we got the highest accuracy of 93.31% using Random Forest.

Download Full-text

Machine Learning Approaches for Predicting Bisphosphonate-Related Osteonecrosis in Women with Osteoporosis Using VEGFA Gene Polymorphisms

Journal of Personalized Medicine ◽

10.3390/jpm11060541 ◽

2021 ◽

Vol 11 (6) ◽

pp. 541

Author(s):

Jin-Woo Kim ◽

Jeong Yee ◽

Sang-Hyeon Oh ◽

Sun-Hyun Kim ◽

Sun-Jong Kim ◽

...

Keyword(s):

Machine Learning ◽

Risk Factors ◽

Logistic Regression ◽

Multivariable Analysis ◽

Significant Risk ◽

Machine Learning Algorithms ◽

Receiver Operating Curve ◽

Support Vector ◽

Learning Approaches ◽

Wild Type

Objective: This nested case–control study aimed to investigate the effects of VEGFA polymorphisms on the development of bisphosphonate-related osteonecrosis of the jaw (BRONJ) in women with osteoporosis. Methods: Eleven single nucleotide polymorphisms (SNPs) of the VEGFA were assessed in a total of 125 patients. Logistic regression was performed for multivariable analysis. Machine learning algorithms, namely, fivefold cross-validated multivariate logistic regression, elastic net, random forest, and support vector machine, were developed to predict risk factors for BRONJ occurrence. Area under the receiver-operating curve (AUROC) analysis was conducted to assess clinical performance. Results: The VEGFA rs881858 was significantly associated with BRONJ development. The odds of BRONJ development were 6.45 times (95% CI, 1.69–24.65) higher among carriers of the wild-type rs881858 allele compared with variant homozygote carriers after adjusting for covariates. Additionally, variant homozygote (GG) carriers of rs10434 had higher odds than those with wild-type allele (OR, 3.16). Age ≥ 65 years (OR, 16.05) and bisphosphonate exposure ≥ 36 months (OR, 3.67) were also significant risk factors for BRONJ occurrence. AUROC values were higher than 0.78 for all machine learning methods employed in this study. Conclusion: Our study showed that the BRONJ occurrence was associated with VEGFA polymorphisms in osteoporotic women.

Download Full-text