Performance Of Machine Learning Models Research Articles

Improving class probability estimates in asymmetric health data classification: An experimental comparison of novel calibration methods

In the context of health data classification, imbalanced and asymmetric class distributions can significantly impact the performance of machine learning models. One critical aspect affected by these issues is the reliability of class probability estimates, which are crucial for informed decision-making in healthcare applications. Instead of predicting class values directly for a classification problem, it can be more convenient to predict the probability of an observation belonging to each possible class. This research aims to address the challenges posed by imbalanced and asymmetric responses in health data classification by evaluating the effectiveness of recent calibration methods in improving class probability estimates. We propose Beta calibration techniques and the Stratified Brier score and Jaccard's Score as novel calibration methods and evaluation metrics respectively. The experimental comparison involves implementing and assessing various calibration techniques to determine their impact on model performance and calibration accuracy of simulated and healthcare datasets with varying imbalance ratios. Our results show that the Beta calibration method consistently improved the classifiers' predictive ability. The findings of this study provide valuable insights into selecting the most suitable calibration method for enhancing class probability estimates in healthcare-related machine learning tasks.

Identification of Power Quality Disturbances in Electrical Distribution System using Fast Fourier Transforms and Super Learner Ensembles

Currently, the use of non-linear loads and equipment, as well as renewable energy sources injected into the power system, tends to increase. As a result, the waveform of the electrical signal changes, and distortion occurs in the distribution system, which affects the quality and reliability of the electrical system. Importantly, sometimes this leads to malfunctions in protection equipment. This paper presents the algorithm for power quality disturbance (PQD) identification in electrical distribution systems, which involves three main steps: (1) Generating simulated waveforms using a signal processing approach; (2) extracting features using the Fast Fourier Transforms (FFT) technique; and (3) identifying the type of PQD using Super Learner Ensembles (SLE), which employs cross-validation to assess the performance of multiple machine learning models. Subsequently, the model’s efficiency is verified and tested using data from electronic energy meters installed in the distribution system of the Provincial Electricity Authority (PEA). The accuracy resulting from synthetic and experimental data sets is 99.90% and 99.69%, respectively. The results indicate that the model performs well in identifying power quality disturbances and achieves high accuracy.

Good results from sensor data: Performance of machine learning algorithms for regression problems in chemical sensors

Accurately predicting unseen data, instead of mere memorization of training examples, is a critical goal of machine learning. This generalization is particularly important in the field of chemical sensors, where the ability to accurately predict the chemical properties or concentration levels of unknown samples is crucial. The paper presents a comprehensive yet accessible introduction to various machine learning concepts, highlighting the importance of model interpretability and generalization in ensuring reliable and accurate results in this context. Nonlinear sensor array data are utilized to introduce key concepts (e.g., bias-variance tradeoff) and techniques (linear models, partial least squares regression, support vector machines, k-nearest neighbors, decision trees, ensemble methods, automated machine learning, symbolic regression, and artificial neural networks), providing a solid foundation to make informed decisions when selecting machine learning techniques for sensor-specific regression applications. The results clearly indicate a number of conclusions. First, overparameterized deep feedforward neural networks show great accuracy and generalization when trained on a sufficiently large dataset. Second, symbolic regression models proved to be more accurate than deep feedforward neural networks and classical machine learning techniques on smaller datasets. Third, the performance of various machine learning models was dataset-dependent, showing the importance of comparative studies to determine the most suitable approach. It is clear that the optimal model cannot be known a priori. This paper aims to provide a starting point for investigations on the performance of different machine learning techniques in chemical sensor applications.

Machine-Learning-Based Path Loss Prediction for Vehicle-to-Vehicle Communication in Highway Environments

Vehicle-to-vehicle (V2V) communication, which plays an important role in intelligent transportation systems, has been statistically proven to improve traffic efficiency and reduce the probability of accidents. In real-world applications, it is critical to accurately estimate the path loss parameter in communication channels due to the variable and complex propagation environments often encountered in inter-vehicle communication scenarios. This paper presents a study on various machine learning methods to improve path loss estimation in V2V communication using a dataset (192,000 observations) obtained from field measurements of highway environments in the Trabzon and Gümüşhane provinces in Türkiye. For this purpose, path loss estimation was carried out with different machine learning algorithms such as Artificial Neural Networks, Random Forest, Linear Regression, Gradient Boosting, Support Vector Regression, and AdaBoost by using various environmental and system features. Then, performance comparisons were conducted between machine learning methods and traditional empirical approaches such as log-distance, two-ray, and log-ray. Examining the outputs reveals that machine learning methods outperform traditional methods and yield results quickly. As a result, the Random Forest and Gradient Boosting methods demonstrated the highest prediction performances, with R2 values of 0.97 and 0.96, MAE values of 0.0557 and 0.0701, and RMSE values of 0.0774 and 0.0964, respectively, outperforming both empirical methods, other machine learning techniques, and the existing studies based on V2V. Overall, our study provides significant contributions to the existing literature by providing a comprehensive parameter set for highway environments, examining the path loss prediction performance of machine learning models with different capabilities, and comparing them with traditional methods. This study not only fills a critical gap in the existing literature but also highlights the necessity, efficiency, and originality of machine learning approaches for improving reliable V2V communication systems.

Combining Multiple Machine Learning Methods Based on CARS Algorithm to Implement Runoff Simulation

Runoff forecasting is crucial for water resource management and flood safety and remains a central research topic in hydrology. Recent advancements in machine learning provide novel approaches for predicting runoff. This study employs the Competitive Adaptive Reweighted Sampling (CARS) algorithm to integrate various machine learning models into a data-driven rainfall–runoff simulation model. We compare the forecasting performance of different machine learning models to improve rainfall–runoff prediction accuracy. This study uses data from the Maduwang hydrological station in the Bahe river basin, which contain 12 measured flood events from 2000 to 2010. Historical runoff and areal mean rainfall serve as model inputs, while flood data at different lead times are used as model outputs. Among the 12 flood events, 9 are used as the training set, 2 as the validation set, and 1 as the testing set. The results indicate that the CARS-based machine learning model effectively forecasts floods in the Bahe River basin. Under the prediction period of 1 to 6 h, the model achieves high forecasting accuracy, with the average NSE ranging from 0.7509 to 0.9671 and the average R2 ranging from 0.8397 to 0.9413, though the accuracy declines to some extent as the lead time increases. The model accurately predicts peak flow and performs well in forecasting high flow and recession flows, though peak flows are somewhat underestimated for longer lead times. Compared to other machine learning models, the SVR model has the highest average RMSE of 0.942 for a 1–6 h prediction period. It exhibits the smallest deviation among low-, medium-, and high-flow curves, with the lowest NRMSE values across training, validation, and test sets, demonstrating better simulation performance and generalization capability. Therefore, the machine learning model based on CARS feature selection can serve as an effective method for flood forecasting. The related findings provide a new forecasting method and scientific decision-making basis for basin flood safety.

A performance and interpretability assessment of machine learning models for rainfall prediction in the Republic of Ireland

Rainfall prediction significantly impacts agriculture, water reserves, and preparations for flooding conditions. This research examines the performance and interpretability of machine learning (ML) models for rainfall prediction in the Republic of Ireland. The study uses a brute force approach and the Leave One Feature Out (LOFO) methodology to evaluate model performance under highly correlated variables. Results reveal consistent performance across ML algorithms, with average Area Under the Curve Precision–Recall (AUC-PR) scores ranging from 0.987 to 1.000, with certain features such as atmospheric pressure and soil moisture deficits demonstrating significant influence on prediction outcomes.SHapley Additive exPlanations (SHAP) values provide insights into feature importance, reaffirming the significance of atmospheric pressure and soil moisture deficits in rainfall prediction. This study underscores the importance of feature selection and interpretability in enhancing the accuracy and usability of ML models for rainfall prediction in Ireland.

Machine Learning for Patient-Based Real-Time Quality Control (PBRTQC), Analytical and Preanalytical Error Detection in Clinical Laboratory.

The rapidly evolving field of machine learning (ML), along with artificial intelligence in a broad sense, is revolutionising many areas of healthcare, including laboratory medicine. The amalgamation of the fields of ML and patient-based real-time quality control (PBRTQC) processes could improve the traditional PBRTQC and error detection algorithms in the laboratory. This narrative review discusses published studies on using ML for the detection of systematic errors, non-systematic errors, and combinations of different types of errors in clinical laboratories. The studies discussed used ML for detecting bias, the requirement for re-calibration, samples contaminated with intravenous fluid or EDTA, delayed sample analysis, wrong-blood-in-tube errors, interference or a combination of different types of errors, by comparing the performance of ML models with human validators or traditional PBRTQC algorithms. Advantages, limitations, the creation of standardised ML models, ethical and regulatory aspects and potential future developments have also been discussed in brief.

Evaluating machine learning models in predicting dam inflow and hydroelectric power production in multi-purpose dams (case study: Mahabad Dam, Iran)

This study aimed to forecast dam inflows and subsequently predict its capability in producing HEPP using machine learning and evolutionary optimization techniques. Mahabad Dam, located in the northwest of Iran and recognized as one of the nation’s key dams, served as a case study. First, artificial neural networks (ANN) and support vector regression (SVR) were employed to predict dam inflows, with optimization of parameters achieved through Harris hawks optimization (HHO), a robust optimization technique. The data of temperature, precipitation, and dam inflow over a 24-year period on a monthly basis, incorporating various lag times, were used to train these machines. Then, HEPP from the dam was predicted using temperature, precipitation, dam inflow, and dam evaporation as input variables. The models were applied to data covering the years 2000 to 2020. The results of the first part indicated both hybrid models (HHO-ANFIS and HHO-SVR) improved the prediction performance compared to the single models. Based on the results of Taylor’s diagram and the error evaluation criteria, the HHO-ANFIS hybrid model (RMSE, MAE, and NSE of 3.90, 2.41, and 0.86, respectively) exerted better performance than HHO-SVR (RMSE, MAE, and NSE of 4.39, 2.70, and 0.86, respectively). The results of the second part showed that using the HHO algorithm to optimize single models (RMSE, MAE, and NSE of 0.2, 10, and 0.90, respectively) predicted HEPP better than single models (RMSE, MAE, and NSE of 0.2, 10, and 0.90, respectively). The results of Taylor’s diagram also showed that the HHO-ANFIS model exerted better performance. The findings of this study indicated the promising performance of machine learning models optimized by metaheuristic algorithms in the simultaneous prediction of dam inflows and HEPP in multi-purpose dams for better management and allocation of surface water resources.

A new understanding of the alkali-silica reaction expansion in concrete using a hybrid ensemble model

Alkali-Silica Reaction (ASR) is a critical issue affecting the durability of concrete structures, leading to significant long-term maintenance costs and safety concerns. This study addresses the need for accurate prediction of ASR-induced expansion in concrete, leveraging advanced machine learning techniques. Two hybrid ensemble methods, the inverse variance method and the artificial neural network-based ensemble method, were proposed to integrate the individual models to enhance the predictive capabilities. A comprehensive experimental database comprising over 1997 sets of ASR expansion data was collected from relevant literature with the environmental conditions, specimen geometry and material composition as the input variables and specimen strain as the output variable. The findings indicate that the predictive performance of individual models falls short of the performance of the hybrid ensemble machine learning model, with the inverse variance-based ensemble model exhibiting the highest performance (R = 97.2 %, RMSE = 0.066 mm). The Shapley Additive Explanations analysis reveals that the most influential factors include reaction days, alkali content, aggregate particle size, and silica content, contributing to up to 35 % of the variation in ASR expansion. This work offers valuable insights for the efficient assessment of durability of concrete structures concerning ASR expansion.

Comparative Analysis of Machine Learning Models and Key Risk Factors in Advancing Predictive Analytics for Coronary Heart Disease Over a Decade

This study aimed to identify key risk factors for coronary heart disease (CHD) and assess the performance of various machine learning models in predicting 10-year CHD risk. We conducted an exploratory data analysis using data from the Framingham Heart Study, which included 4,238 participants and 15 potential risk factors, to understand the distribution of variables and relationships. We used Chi-square and Mann-Whitney U tests to identify significant associations between risk factors and coronary heart disease. Logistic regression, random forest and support vector machine (SVM) models were established and their prediction accuracy and area under receiver operating characteristic curve (AUC) were evaluated. The results showed that age, systolic blood pressure and history of hypertension were the most influential risk factors. Logistic regression accuracyand AUCwere the highest, better than random forest and SVM. This indicates that we can pay more attention to such factors as age, systolic blood pressure and history of hypertension in subsequent CHD studies, and mainly use logistic regression model to predict coronary heart disease and optimize it.

Comparison of Machine Learning Models for Predicting Interstitial Glucose Using Smart Watch and Food Log

This study examines the performance of various machine learning (ML) models in predicting Interstitial Glucose (IG) levels using data from wrist-worn wearable sensors. The insights from these predictions can aid in understanding metabolic syndromes and disease states. A public dataset comprising information from the Empatica E4 smart watch, the Dexcom Continuous Glucose Monitor (CGM) measuring IG, and a food log was utilized. The raw data were processed into features, which were then used to train different ML models. This study evaluates the performance of decision tree (DT), support vector machine (SVM), Random Forest (RF), Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN), Gaussian Naïve Bayes (GNB), lasso cross-validation (LassoCV), Ridge, Elastic Net, and XGBoost models. For classification, IG labels were categorized into high, standard, and low, and the performance of the ML models was assessed using accuracy (40–78%), precision (41–78%), recall (39–77%), F1-score (0.31–0.77), and receiver operating characteristic (ROC) curves. Regression models predicting IG values were evaluated based on R-squared values (−7.84–0.84), mean absolute error (5.54–60.84 mg/dL), root mean square error (9.04–68.07 mg/dL), and visual methods like residual and QQ plots. To assess whether the differences between models were statistically significant, the Friedman test was carried out and was interpreted using the Nemenyi post hoc test. Tree-based models, particularly RF and DT, demonstrated superior accuracy for classification tasks in comparison to other models. For regression, the RF model achieved the lowest RMSE of 9.04 mg/dL with an R-squared value of 0.84, while the GNB model performed the worst, with an RMSE of 68.07 mg/dL. A SHAP analysis identified time from midnight as the most significant predictor. Partial dependence plots revealed complex feature interactions in the RF model, contrasting with the simpler interactions captured by LDA.

A comprehensive study of pharmaceutics solubility in supercritical solvent through diverse thermodynamic and hybrid Machine learning approaches

The pharmaceutical industry is increasingly drawn to the research of innovative drug delivery systems through the use of supercritical CO2 (scCO2)-based techniques. Measuring the solubility of drugs in scCO2 at varying conditions is a crucial parameter in this context. In this research, the supercritical solubility of two pharmaceutical ingredients, namely Febuxostat and Chlorpromazine, has been assessed theoretically using various thermodynamic approaches, including PR, SRK, UNIQUAC, and Wilson models. Additionally, hybrid machine learning models of PO-GPR, and PO-KNN were applied to anticipate the supercritical solubility of these medicines. Verification of the accuracy of each model for each pharmaceutical substance is conducted against previously reported experimental solubility data. In the comparison between the SRK and PR models, it is observed that the SRK model displays greater precision in correlating the solubility of both drugs. It consistently achieves a mean Radj value of 0.995 across all cases and mean AARD% values of 14.47 and 9.30 for Febuxostat and Chlorpromazine, respectively. Furthermore, the findings indicate that the UNIQUAC model surpasses the Wilson model in precisely representing the solubility of both medicines. It consistently achieves a mean Radj value higher than 0.985 across both cases and mean AARD% values of 11.39 and 7.08 for Febuxostat and Chlorpromazine, respectively. Additionally, the performance of both hybrid machine learning models proved to be excellent in anticipating the supercritical solubility of both compounds.

Sand Production Prediction with Machine Learning using Input Variables from Geological and Operational Conditions in the Karazhanbas Oilfield, Kazakhstan

AbstractThis paper describes a comprehensive approach to predict sand production in the Karazhanbas oilfield using machine learning (ML) techniques. By analyzing data from 2000 wells, the research uncovered the complex dynamics of sand production and emphasized the critical need for accurately predicting the peak sand mass and its occurrence time. ML techniques can have a significant impact on prediction of sand production and on the optimization of oilfield operation, which can be improved with the combined use of enriched training data and domain-specific knowledge. The research underscored the influence of geological factors, especially fault proximity, on prediction accuracy. Domain and field knowledge is needed to formulate different production scenarios for prediction purposes such that the relevant data can be selected for the training of ML models. Moreover, new metrics are needed to evaluate model performance as the applied method is tailored for different operational strategies. As the peak sand mass is considered a pivotal event in field operation, new metrics in terms of peak prediction accuracy and peak time prediction accuracy were introduced to evaluate the performance of ML models. A suite of ML algorithms was employed in the study, which demonstrated notable accuracy in the classification of sand-producing wells.

Predicting 30-day mortality in severely injured elderly patients with trauma in Korea using machine learning algorithms: a retrospective study.

The number of elderly patients with trauma is increasing; therefore, precise models are necessary to estimate the mortality risk of elderly patients with trauma for informed clinical decision-making. This study aimed to develop machine learning based predictive models that predict 30-day mortality in severely injured elderly patients with trauma and to compare the predictive performance of various machine learning models. This study targeted patients aged ≥65 years with an Injury Severity Score of ≥15 who visited the regional trauma center at Chungbuk National University Hospital between 2016 and 2022. Four machine learning models-logistic regression, decision tree, random forest, and eXtreme Gradient Boosting (XGBoost)-were developed to predict 30-day mortality. The models' performance was compared using metrics such as area under the receiver operating characteristic curve (AUC), accuracy, precision, recall, specificity, F1 score, as well as Shapley Additive Explanations (SHAP) values and learning curves. The performance evaluation of the machine learning models for predicting mortality in severely injured elderly patients with trauma showed AUC values for logistic regression, decision tree, random forest, and XGBoost of 0.938, 0.863, 0.919, and 0.934, respectively. Among the four models, XGBoost demonstrated superior accuracy, precision, recall, specificity, and F1 score of 0.91, 0.72, 0.86, 0.92, and 0.78, respectively. Analysis of important features of XGBoost using SHAP revealed associations such as a high Glasgow Coma Scale negatively impacting mortality probability, while higher counts of transfused red blood cells were positively correlated with mortality probability. The learning curves indicated increased generalization and robustness as training examples increased. We showed that machine learning models, especially XGBoost, can be used to predict 30-day mortality in severely injured elderly patients with trauma. Prognostic tools utilizing these models are helpful for physicians to evaluate the risk of mortality in elderly patients with severe trauma.

Forecasting the architecture billings index (ABI) using machine learning predictive models

PurposeThis research aims to forecast the ABI as a leading indicator of U.S. construction activities, applying multivariate machine learning predictive models over different horizons and utilizing the nonlinear and long-term dependencies between the ABI and macroeconomic and construction market variables. To assess the applicability of the machine learning models, six multivariate machine learning predictive models were developed considering the relationships between the ABI and other construction market and macroeconomic variables. The forecasting performances of the developed predictive models were evaluated in different forecasting scenarios, such as short-term, medium-term, and long-term horizons comparable to the actual timelines of construction projects.Design/methodology/approachThe architecture billings index (ABI) as a macroeconomic indicator is published monthly by the American Institute of Architects (AIA) to evaluate business conditions and track construction market movements. The current research developed multivariate machine learning models to forecast ABI data for different time horizons. Different macroeconomic and construction market variables, including Gross Domestic Product (GDP), Total Nonresidential Construction Spending, Project Inquiries, and Design Contracts data were considered for predicting future ABI values. The forecasting accuracies of the machine learning models were validated and compared using the short-term (one-year-ahead), medium-term (three-year-ahead), and long-term (five-year-ahead) ABI testing datasets.FindingsThe experimental results show that Long Short Term Memory (LSTM) provides the highest accuracy among the machine learning and traditional time-series forecasting models such as Vector Error Correction Model (VECM) or seasonal ARIMA in forecasting the ABIs over all the forecasting horizons. This is because of the strengths of LSTM for forecasting temporal time series by solving vanishing or exploding gradient problems and learning long-term dependencies in sequential ABI time series. The findings of this research highlight the applicability of machine learning predictive models for forecasting the ABI as a leading indicator of construction activities, business conditions, and market movements.Practical implicationsThe architecture, engineering, and construction (AEC) industry practitioners, investment groups, media outlets, and business leaders refer to ABI as a macroeconomic indicator to evaluate business conditions and track construction market movements. It is crucial to forecast the ABI accurately for strategic planning and preemptive risk management in fluctuating AEC business cycles. For example, cost estimators and engineers who forecast the ABI to predict future demand for architectural services and construction activities can prepare and price their bids more strategically to avoid a bid loss or profit loss.Originality/valueThe ABI data have been forecasted and modeled using linear time series models. However, linear time series models often fail to capture nonlinear patterns, interactions, and dependencies among variables, which can be handled by machine learning models in a more flexible manner. Despite the strength of machine learning models to capture nonlinear patterns and relationships between variables, the applicability and forecasting performance of multivariate machine learning models have not been investigated for ABI forecasting problems. This research first attempted to forecast ABI data for different time horizons using multivariate machine learning predictive models using different macroeconomic and construction market variables.

Establishing reflex test rules for platelet fluorescent counting method using machine learning models on Sysmex XN-series hematology analyzer.

The platelet fluorescent counting (PLT-F) method is utilized as a reflex test method following the initial test of the platelet impedance counting (PLT-I) method in clinical practice on the Sysmex XN-series automated hematology analyzer. Our aim is to establish reflex test rules for the PLT-F method by combining multiple parameters provided by the "CBC + DIFF" mode of the Sysmex XN-series automated hematology analyzer. We tested 120 samples to evaluate the baseline bias between the PLT-F and PLT-I methods. Then, we selected 1256 samples to establish and test reflex test rules using seven machine learning models (decision Tree, random forest, neural network, logistic regression, k-nearest neighbor, support vector machine, and Naive Bayes). The training set and test set were divided at a ratio of 7:3. We evaluated the performance of machine learning models on the test set using various metrics to select the most valuable model. The PLT-F method exhibited a high degree of correlation with the PLT-I method (r = 0.998). The random forest model emerged as the most valuable, boasting an accuracy of 0.893, an area under the curve of 0.954, an F1 score of 0.771, a recall of 0.719, a precision of 0.831, and a specificity of 0.950. The most important variable in the random forest model was mean cell volume, weighted at 15.09%. The random forest model, which demonstrated high efficiency in our study, can be used to establish PLT reflex test rules based on the PLT-F method for the Sysmex XN-series automated hematology analyzer.

Combining Machine Learning techniques and Genetic Algorithm for predicting run times of High Performance Computing jobs

This study proposes a novel approach combining Machine Learning (ML) techniques and Genetic Algorithms (GA) for predicting High-Performance Computing (HPC) job run times. The objective is to create a prediction method universally applicable to any HPC system, irrespective of workload characteristics, application specific parameters, user behavior, or hardware architecture. Since user-supplied run time estimates are often inaccurate, we aim to categorize job runtimes into several classes, allowing users to select appropriate classes for their jobs. A Genetic Algorithm is developed to optimally define these runtime classes, determining both the number of classes and the time intervals represented by them. Four Machine Learning algorithms (K Nearest Neighbours, Support Vector Regression, Extreme Gradient Boosting and Deep Neural Networks) are implemented for run time prediction. A unique set of features extracted from historical job data serves as input to the Machine Learning models. The generalized nature of our method is demonstrated by validating its performance on data from six clusters with distinct configurations, applications and runtime distributions. Our results illustrate the superior performance of Machine Learning models incorporating GA-defined runtime classes for all datasets. Across all six datasets, our method achieves R2 scores exceeding 0.8, and accuracy greater than 0.7.

Improving Low‐Cloud Fraction Prediction Through Machine Learning

AbstractIn this study, we evaluated the performance of machine learning (ML) models (XGBoost) in predicting low‐cloud fraction (LCF), compared to two generations of the community atmospheric model (CAM5 and CAM6) and ERA5 reanalysis data, each having a different cloud scheme. ML models show a substantial enhancement in predicting LCF regarding root mean squared errors and correlation coefficients. The good performance is consistent across the full spectrums of atmospheric stability and large‐scale vertical velocity. Employing an explainable ML approach, we revealed the importance of including the amount of available moisture in ML models for representing spatiotemporal variations in LCF in the midlatitudes. Also, ML models demonstrated marked improvement in capturing the LCF variations during the stratocumulus‐to‐cumulus transition (SCT). This study suggests ML models' great potential to address the longstanding issues of “too few” low clouds and “too rapid” SCT in global climate models.

An Assessment of Contemporary Methods and Data-Enabled Approaches for Early Cataract Detection

Background: Cataracts are common causes of visual impairment. Preventing blindness requires an early and accurate diagnosis. This review examines current cataract diagnosis strategies, explores data-driven machine learning algorithms for early detection, investigates the use of artificial intelligence (AI) approaches, assesses improvements in cataract detection accuracy, identifies research gaps, and provides recommendations for future studies. Methods: We gathered labelled cataract and non-cataract fundus data from the Kaggle. Scholarly publications were sourced from reliable databases such as ProQuest, IEEE, ELSEVIER, Google Scholar, and PubMed. A detailed literature search with specific terms expanded the scope of this review. We included studies that used cataract and non-cataract fundus eye images from cross-sectional, retrospective, and prospective studies. The quality assessment used the AMSTAR tool, considering factors such as literature search comprehensiveness, study selection criteria, data extraction methodologies, and study validity (Table 1). Results: This study encompassed 130 research publications, focusing on machine learning models and clinical-based diagnostic approaches for early-stage cataract identification. The performance of machine-learning models is influenced by factors such as dataset noise and limited reliable data. Barriers to the successful implementation of AI for cataract diagnosis were identified. Conclusions: This review emphasises the obstacles hindering the broad application of AI in cataract diagnosis. Addressing these findings is vital for developing strategies to overcome these challenges and enhance cataract detection systems. To achieve improved accuracy and efficiency in cataract diagnosis, future research should prioritise efforts to enhance dataset availability and quality, reduce data noise, and refine machine-learning algorithms. Unlocking the full potential of AI and/or machine learning can lead to significant breakthroughs in cataract diagnosis, ultimately resulting in better patient outcomes and reduced visual impairments.

Improving Kui digit recognition through machine learning and data augmentation techniques

Speech digit recognition research is growing decisively, and a bulk of digit recognition algorithms are used in European and a few Asian languages. Kui is a low-resourced tribal language locally used in several states of India. Despite its significance, there is not much research on Kui's speech. This research aims to present an in-depth analysis of novel Kui digit recognition using predefined machine learning (ML) techniques. For this purpose, we first gathered spoken numbers i.e. from 0 to 9 of eight different speakers containing a total of 200 words. Secondly, we choose the numbers: ଶୂନ (zero), ଏକ (one), ଦୁଇ (two), ତିନି(three), ସାରି(four), ପାସ (five), ସଅ (six), ସାତ (seven), ଆଟ (eight), ନଅ (nine). Meanwhile, we build nine different ML models to recognize Kui digits that take the Mel-frequency cepstral coefficients (MFCCs) method to extract the relevant features for model predictions. Finally, we compared the performance of ML models for both augmented and non-augmented Kui data. The result shows that the SVM+Augmentation method for Kui digit recognition combined obtained the highest accuracy of 83% than other methods. Moreover, the difficulties and potential prospects for Kui digit recognition are also highlighted in this work.