Article
Computer Science, Information Systems
Nasir Jalal, Arif Mehmood, Gyu Sang Choi, Imran Ashraf
Summary: This study presents an improved random forest model for text classification, called IRFTC, which incorporates bootstrapping and random subspace methods. It achieves better classification performance and outperforms other machine learning models.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Mathematics, Applied
Gabriel Aguilera-Venegas, Amador Lopez-Molina, Gemma Rojo-Martinez, Jose Luis Galan-Garcia
Summary: The main goals of this work are to study and compare machine learning algorithms in predicting the development of type 2 diabetes mellitus. Four classification algorithms, including Decision Tree, Random Forest, kNN, and Neural Networks, were examined and compared for their accuracy in predicting the incidence of type 2 diabetes mellitus seven and a half years in advance. The study not only compared the techniques, but also fine-tuned the hyperparameters of each algorithm. The algorithms were implemented using R language and the data base used was obtained from the nation-wide cohort di@bet.es study. This work provides the accuracy of each algorithm and identifies the best technique and hyperparameters for this problem.
JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS
(2023)
Article
Clinical Neurology
Rasmus Bach Nedergaard, Matthew Scott, Anne-Marie Wegeberg, Tina Okdahl, Joachim Starling, Birgitte Brock, Asbjarn Mohr Drewes, Christina Brock
Summary: This study used supervised machine learning to classify the severity of cardiovascular autonomic neuropathy (CAN). The results showed that beat-to-beat measures, inflammation markers, disease-duration, and age were the most important features for characterizing CAN. It was suggested that monitoring cardiac reflex responses closely and targeting systemic low-grade inflammation could help diagnose and prevent the development of CAN.
CLINICAL NEUROPHYSIOLOGY
(2023)
Article
Biochemical Research Methods
Karlo Abnoosian, Rahman Farnoosh, Mohammad Hassan Behzadi
Summary: In this study, we propose a pipeline-based multi-classification framework to predict diabetes using an imbalanced dataset of Iraqi diabetic patients. The framework addresses challenges such as limited labeled data, frequent missing values, and dataset imbalance by incorporating various pre-processing techniques and implementing multiple machine learning models. The proposed model outperforms other models in predicting diabetes and achieves high accuracy, precision, recall, F1-score, and AUC values.
BMC BIOINFORMATICS
(2023)
Article
Computer Science, Information Systems
Md Nasim Adnan, Ryan H. L. Ip, Michael Bewong, Md Zahidul Islam
Summary: The proposed decision forest algorithm in this paper achieves better balance through effective synchronization of diversity from different sources, leading to significant improvement in accuracy according to empirical evaluations. It is also competitive in terms of complexity and other relevant parameters.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Mohammad Aktaruzzaman Khan, Sayed Allamah Iqbal, Maliha Sanjida Khan, Md. Golam Hafez
Summary: This study utilizes machine learning and factor analysis to predict job satisfaction. Despite the limitations of factor analysis, machine learning algorithms can overcome these challenges. The study finds that management support, equity, non-financial compensation, and financial compensation are highly effective in predicting job satisfaction.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2023)
Article
Public, Environmental & Occupational Health
Michael Onyema Edeh, Osamah Ibrahim Khalaf, Carlos Andres Tavera, Sofiane Tayeb, Samir Ghouali, Ghaida Muttashar Abdulsahib, Nneka Ernestina Richard-Nnabu, AbdRahmane Louni
Summary: A model that accurately predicts the likelihood of developing diabetes was developed using four machine learning classification algorithms. The experiments on different databases showed that the random forest and SVM algorithms had the highest accuracy.
FRONTIERS IN PUBLIC HEALTH
(2022)
Article
Environmental Studies
Patricia Bittencourt Tavares das Neves, Claudio Jose Cavalcante Blanco, Andre Augusto Azevedo Montenegro Duarte, Filipe Bittencourt Souza das Neves, Isabela Bittencourt Souza das Neves, Marcelo Henrique de Paula dos Santos
Summary: This study investigates the impact of clandestine and regular road networks on Amazon forest deforestation and uses a time series analysis to predict deforestation trends over the next three decades. The results suggest that deforestation will continue to intensify, albeit with decreasing rates of forest loss over a ten-year period.
Article
Biochemistry & Molecular Biology
Ulan Tore, Aibek Abilgazym, Angel Asunsolo-del-Barco, Milan Terzic, Yerden Yemenkhan, Amin Zollanvari, Antonio Sarria-Santamera
Summary: Endometriosis is a challenging and enigmatic disease associated with multiple conditions. Machine learning can be used to improve its diagnosis. A predictive model using logistic regression, decision tree, random forest, AdaBoost, and XGBoost achieved a satisfactory performance but more research is needed to enhance its accuracy.
Article
Computer Science, Information Systems
Ashima Singh, Arwinder Dhillon, Neeraj Kumar, M. Shamim Hossain, Ghulam Muhammad, Manoj Kumar
Summary: This article introduces an ensemble-based framework called eDiaPredict, which utilizes different machine learning algorithms to predict diabetes status among patients with high accuracy. The use of modern computational intelligence in healthcare, particularly in predicting disease onset and recurrence, identifying survival analysis biomarkers, etc., highlights the importance of machine learning techniques in the medical field.
ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS
(2021)
Article
Biology
Md Mamun Ali, Bikash Kumar Paul, Kawsar Ahmed, Francis M. Bui, Julian M. W. Quinn, Mohammad Ali Moni
Summary: The study aimed to identify machine learning classifiers with the highest accuracy for heart disease diagnostic purposes; several supervised machine-learning algorithms were applied and compared, with random forests algorithm achieving the best performance; feature importance scores were used to find features with high predictive power for heart disease.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Medical Informatics
Alan Brnabic, Lisa M. Hess
Summary: This study conducted a systematic literature review on the application of machine learning in informing decision making at the patient-provider level. It found that there is a wide variety of methods and validation strategies used in current studies, indicating the need for further improvement, particularly in ensuring decisions are based on high-quality evidence.
BMC MEDICAL INFORMATICS AND DECISION MAKING
(2021)
Article
Engineering, Environmental
Jyoti Singh, Sarvanshi Swaroop, Vishal Mishra
Summary: Researchers used decision tree and random forest regression algorithms to model and predict the adsorption capacity of fired and non-fired beads. The results showed that the random forest regression algorithm had better performance. This approach is an effective way to combat heavy metal contamination while reducing the number of experiments required.
JOURNAL OF ENVIRONMENTAL CHEMICAL ENGINEERING
(2022)
Article
Energy & Fuels
Amer Alanazi, Ahmed Farid Ibrahim, Saleh Bawazer, Salaheldin Elkatatny, Hussein Hoteit
Summary: For the purpose of carbon capture, utilization, and storage, this study presents a machine-learning framework that predicts CO2 adsorption in coal formations based on various coal properties and testing conditions. The ML techniques used include decision tree regression, random forests, gradient boost regression, K-nearest neighbor, artificial neural network, function network, and adaptive neuro-fuzzy inference system.
INTERNATIONAL JOURNAL OF COAL GEOLOGY
(2023)
Article
Ecology
Yoshifumi Masago, Maychee Lian
Summary: Climate change affects the flowering dates of Yoshino cherry trees, and models were developed using machine learning algorithms and climate data to estimate these dates. The analysis suggests that low winter temperatures and high spring temperatures advance the flowering dates.
ECOLOGICAL INFORMATICS
(2022)