PIMA Dataset Research Articles

Explainable deep learning for diabetes diagnosis with DeepNetX2

Diabetes is a leading health global health challenge because of its high blood sugar levels and the risk of extensive damage to other internal organs. Early and accurate identification of diabetes is important because it may cause other diseases including heart diseases and nerve damage. Despite the success of using machine learning, especially deep learning in automated diabetes diagnosis. These models are mostly black boxes which rarely offer comprehensive explanations and interpretations of the results. This study introduces DeepNetX2, a proposed custom deep neural network designed to overcome these challenges by integrating Explainable Artificial Intelligence (XAI) techniques, specifically Local Interpretable Model-agnostic Explanations (LIME) and Shapley Additive Explanations (SHAP). These techniques make the decision-making process of the model transparent, thereby increasing the credibility of the predictions. The proposed methodology entails a comprehensive data preprocessing technique that involves a customized Spearman’s correlation coefficient feature selection strategy. This preprocessing restricts complexity to only relevant features that promote effectiveness, instead of oversimplification, to the point of decreasing efficiency. DeepNetX2 was rigorously tested on three datasets: the PIMA dataset, local private dataset, and Type-2 diabetes dataset, achieving test accuracies of 94.81%, 97.87%, and 97.50%, respectively. These results demonstrate not only the superior performance of DeepNetX2 compared with existing models, but also its enhanced interpretability. Based on the proposed model, an appropriate and rapid strategy for diabetes prediction was developed which is very useful for improving diagnostic integrity and patient health.

Optimized Forest Framework with A Binary Multineighborhood Artificial Bee Colony for Enhanced Diabetes Mellitus Detection

Diabetes mellitus (DM) is a common chronic condition that mainly affects older adults. It's important to identify it early to prevent complications. Machine learning is essential for early detection of DM. This article introduces a new method for detecting DM using a random forest ensemble within an optimized framework. The optimized forest framework depends on finding the best DM features, which are identified using the binary multineighborhood artificial bee colony (BMNABC) technique. During preprocessing, the BMNABC algorithm efficiently identifies important features and then inputs them into the random forest within the optimized forest framework for accurate classification. Five modern DM datasets were used to validate the suggested model. The comparison of the proposed model with other leading models revealed significant insights. The BMNABC + ODF(RFE) model demonstrated exceptional proficiency in detecting diabetes mellitus (DM) across various datasets. It achieved an accuracy of 96.36% and a sensitivity of 99.95% on the merged dataset (130 US and PIMA images). The Iranian Ministry of Health dataset showed an accuracy of 97.28% and a sensitivity of 97.12%. In the Sylhet Diabetes Hospital dataset, the accuracy and sensitivity were 96.81% and 98.07% respectively. However, on the PIMA dataset, the model displayed a nuanced performance, with an accuracy of 77.21% and a sensitivity of 68.83%. Lastly, on the questionnaire dataset, the BMNABC + ODF(RFE) model achieved an accuracy of 96.43% and a sensitivity of 97.15%. These findings emphasize the model's ability to adapt and perform effectively in different clinical environments, outperforming other models in terms of accuracy and sensitivity in detecting DM.

A blockchain-machine learning ecosystem for IoT-Based remote health monitoring of diabetic patients

Diabetes poses a global health challenge, demanding continuous monitoring and expert care for effective management. Conventional monitoring methods lack real-time insights and secure data-sharing capabilities, necessitating innovative solutions that leverage emerging technologies. Existing centralized monitoring systems often entail risks such as data breaches and single points of failure, emphasizing the necessity for a secure, decentralized approach that integrates the Internet of Things (IoT), blockchain, and machine learning for efficient and secure diabetes management. This paper introduces a decentralized, blockchain-based framework for remote diabetes monitoring, IoT sensors, machine learning models, and decentralized applications (DApps). The proposed framework comprises five layers: the IoT Sensor Layer, which collects real-time health data from patients; the Blockchain Layer, leveraging smart contracts on the Ethereum blockchain for secure data sharing and transactions; the machine learning Layer, analyzing patient data to detect diabetes; and the DApps Layer, facilitating interactions between patients, doctors, and hospitals. For intelligent decision-making regarding diabetes based on data collected from different sensors, nine machine learning algorithms, including logistic regression, K-nearest neighbors (KNN), support vector machine (SVM), Decision Tree, Random Forest, AdaBoost, stochastic gradient boosting (SGD), and Naive Bayes, were trained and tested on the PIMA dataset. Based on the performance evaluation parameters such as accuracy, recall, F1-score, and the area under the curve (AUC), it was found that the AdaBoost model achieved the highest predictive accuracy of 92.64%, followed by the Decision Tree with an accuracy of 92.21% in diabetes classification.

Advancing Precision Healthcare Analytics: Machine Learning Approach-es for Diabetes Prognosis using the PIMA Indian Dataset

The prevalence of diabetes presents a significant global health challenge, necessitating practical prognostic tools for timely intervention. Harnessing machine learning offers a promising avenue for accurate prediction, aiding in early detection and prevention. This study delves into the development of machine learning models for diabetes prognosis, leveraging the PIMA Indian dataset. Emphasizing the importance of early detection, the research underscores diabetes as a modifiable condition through lifestyle adjustments. By analyzing diverse healthcare data, including electronic health records and imagery, machine learning algorithms can unveil latent patterns crucial for timely diagnosis. This project has three objectives: 1) To find the significant features contributing to diabetes, 2) To develop a predictive model for diabetes prediction, and 3) To find an optimal model for diabetes prediction, comparing the performance of models. The research methodology encompasses a range of statistical and machine-learning algorithms applied to the PIMA dataset to identify optimal prediction methods. Our research underscores the pivotal role of glucose levels as the foremost predictor of diabetes across all models examined, reaffirming the significance of this metabolic marker in diabetes prognosis. The findings highlight logistic regression efficacy in feature extraction and prediction accuracy over random forest, K nearest neighbors (KNN), and deep neural (DL) networks, suggesting its potential for automated prognostic tools. Furthermore, omics data integration promises to enhance DL model performance, paving the way for robust diagnostic solutions.

SMOTE-based Deep LSTM System with GridSearchCV Optimization for Intelligent Diabetes Diagnosis

Diabetes is a metabolic illness initiated by either inadequate insulin creation by the pancreas or the body's reduced responsiveness to insulin. It is characterised by consistently high levels of blood sugar and symptoms such as frequent urination, thirst, and increased appetite. Untreated diabetes can result in significant problems that impact crucial organs, presenting potentially fatal dangers. In order to address the need for accurate diagnosis, researchers have utilised artificial intelligence to develop the G-LSTM system. This system employs a novel method that combines SMOTE-based deep LSTM and GridSearchCV optimization to classify diabetes. This technique effectively tackles the issue of class imbalance in diabetes datasets, demonstrating an exceptional level of prediction accuracy. When tested on the PIMA dataset, G-LSTM demonstrated exceptional performance with an accuracy of 97.12%. Additionally, it produced high precision, recall, F1-score, AUC, and MCC values of 97.12%, 0.963, 0.954, 0.887, 0.989, and 0.882, respectively. The results highlight the higher performance of the G-LSTM method compared to other techniques, suggesting its use for clinical investigation of diabetes patients. This innovative intelligent diagnostic framework not only demonstrates the potential of artificial intelligence in healthcare, but also highlights its crucial role in enhancing the precision and effectiveness of diabetes diagnosis and treatment.

An intelligent diabetes classification and perception framework based on ensemble and deep learning method.

Sugar in the blood can harm individuals and their vital organs, potentially leading to blindness, renal illness, as well as kidney and heart diseases. Globally, diabetic patients face an average annual mortality rate of 38%. This study employs Chi-square, mutual information, and sequential feature selection (SFS) to choose features for training multiple classifiers. These classifiers include an artificial neural network (ANN), a random forest (RF), a gradient boosting (GB) algorithm, Tab-Net, and a support vector machine (SVM). The goal is to predict the onset of diabetes at an earlier age. The classifier, developed based on the selected features, aims to enable early diagnosis of diabetes. The PIMA and early-risk diabetes datasets serve as test subjects for the developed system. The feature selection technique is then applied to focus on the most important and relevant features for model training. The experiment findings conclude that the ANN exhibited a spectacular performance in terms of accuracy on the PIMA dataset, achieving a remarkable accuracy rate of 99.35%. The second experiment, conducted on the early diabetes risk dataset using selected features, revealed that RF achieved an accuracy of 99.36%. Based on our experimental results, it can be concluded that our suggested method significantly outperformed baseline machine learning algorithms already employed for diabetes prediction on both datasets.

An efficient prediction system for diabetes disease based on machine learning algorithms

Diabetes is a persistent medical condition that arises when the pancreas loses its ability to produce insulin or when the body is unable to utilize the insulin it generates effectively. In today's world, diabetes stands as one of the most prevalent and, unfortunately, one of the deadliest diseases due to certain complications. Timely detection of diabetes plays a crucial role in facilitating its treatment and preventing the disease from advancing further. In this study, we have developed a diabetes prediction model by leveraging a variety of machine learning classification algorithms, including K-Nearest Neighbors (KNN), Naive Bayes, Support Vector Machine (SVM), Decision Tree, Random Forest, and Logistic Regression, to determine which algorithm yields the most accurate predictive outcomes. we employed the famous PIMA Indians Diabetes dataset, comprising 768 instances with nine distinct feature attributes. The primary objective of this dataset is to ascertain whether a patient has diabetes based on specific diagnostic metrics included in the collection. In the process of preparing the data for analysis, we implemented a series of preprocessing steps. The evaluation of performance metrics in this study encompassed accuracy, precision, recall, and the F1 score. The results from our experiments indicate that the K-nearest neighbors’ algorithm (KNN) surpasses other algorithms in effectively differentiating between individuals with diabetes and those without in the PIMA dataset.

Predictive machine learning applying cross industry standard process for data mining for the diagnosis of diabetes mellitus type 2

<p>Currently, type 2 diabetes mellitus is one of the world's most prevalent<br />diseases and has claimed millions of people's lives. The present research aims<br />to know the impact of the use of machine learning in the diagnostic process<br />of type 2 diabetes mellitus and to offer a tool that facilitates the diagnosis of<br />the dis-ease quickly and easily. Different machine learning models were<br />designed and compared, being random forest was the algorithm that generated<br />the model with the best performance (90.43% accuracy), which was integrated<br />into a web platform, working with the PIMA dataset, which was validated by<br />specialists from the Peruvian League for the Fight against Diabetes<br />organization. The result was a decrease of (A) 88.28% in the information<br />collection time, (B) 99.99% in the diagnosis time, (C) 44.42% in the diagnosis<br />cost, and (D) 100% in the level of difficulty, concluding that the application<br />of machine learning can significantly optimize the diagnostic process of type<br />2 diabetes mellitus.</p>

Optimizing diabetes classification with a machine learning-based framework

BackgroundDiabetes is a metabolic disorder usually caused by insufficient secretion of insulin from the pancreas or insensitivity of cells to insulin, resulting in long-term elevated blood sugar levels in patients. Patients usually present with frequent urination, thirst, and hunger. If left untreated, it can lead to various complications that can affect essential organs and even endanger life. Therefore, developing an intelligent diagnosis framework for diabetes is necessary.ResultThis paper proposes a machine learning-based diabetes classification framework machine learning optimized GAN. The framework encompasses several methodological approaches to address the diverse challenges encountered during the analysis. These approaches encompass the implementation of the mean and median joint filling method for handling missing values, the application of the cap method for outlier processing, and the utilization of SMOTEENN to mitigate sample imbalance. Additionally, the framework incorporates the employment of the proposed Diabetes Classification Model based on Generative Adversarial Network and employs logistic regression for detailed feature analysis. The effectiveness of the framework is evaluated using both the PIMA dataset and the diabetes dataset obtained from the GEO database. The experimental findings showcase our model achieved exceptional results, including a binary classification accuracy of 96.27%, tertiary classification accuracy of 99.31%, precision and f1 score of 0.9698, recall of 0.9698, and an AUC of 0.9702.ConclusionThe experimental results show that the framework proposed in this paper can accurately classify diabetes and provide new ideas for intelligent diagnosis of diabetes.

Penerapan Particle Swarm Optimization Untuk Meningkatkan Kinerja Algoritma K-Nearest Neighbor Dalam Klasifikasi Penyakit Diabetes

The study aims to apply the KNN algorithm to classify diabetes by combining the PSO algorithm as a selection so as to obtain the best accuracy from the KNN algorithm in classifying, so that it can be applied to diagnose diabetes. This research consists of several stages of research including dataset collection, data pre-processing, data sharing, finding the optimal k-value to the classification process and performance or accuracy testing. From this research, the accuracy of the KNN algorithm before feature selection using PSO is 75% at k-optimal 19 and after feature selection using PSO is obtained an increase in accuracy to 77.213% at the same k value with features that affect the pima dataset, namely Glucose, Blood Pressure, Skin Thickness, Insulin, BMI, Diabetes Pedigree Function, and Age. Thus the use of KNN with PSO feature selection can be used to identify diabetes because it has a much better level of accuracy.

Research on using AdaBoost with K-Means and SMOTE to predict the incidence of diabetes

Diabetes is one of the most diseases in the world. In the last 40 years, the number of persons worldwide with diabetes has tripled. There were 108 million patients over the age of 18 in 1980 and 422 million in 2014, accounting for 8.5% of the entire population at that time. Diabetes directly caused 1.5 million fatalities worldwide in 2012, with hyperglycemia-related illnesses accounting for 2.2 million deaths. Diabetes is expected to be the 7th greatest cause of death by 2030 according to the World Health Organization. As the risk of diabetes increases, machine learning algorithms are used to improve early diagnosis of diabetes, and various researchers have also done some corresponding algorithms for predicting diabetes machine learning. As a commonly used machine learning algorithm, AdaBoost integrated learning algorithm is superior in the diagnosis and prediction of diabetes mellitus. In this paper, it is proposed that a hybrid model to detect the risk of diabetes. This hybrid model is detected and eliminated by K-means-based outliers, synthesizing the distribution of minority data oversampling techniques (SMOTE), and Adaboost to classify diabetes. According to the final experimental result, the model prediction accuracy is 0.950 after using the hybrid model in the PIMA dataset. In the future, if a larger number of sample training data are utilized for training, the model's accuracy will improve.

A Novel Proposal for Deep Learning-Based Diabetes Prediction: Converting Clinical Data to Image Data.

Diabetes, one of the most common diseases worldwide, has become an increasingly global threat to humans in recent years. However, early detection of diabetes greatly inhibits the progression of the disease. This study proposes a new method based on deep learning for the early detection of diabetes. Like many other medical data, the PIMA dataset used in the study contains only numerical values. In this sense, the application of popular convolutional neural network (CNN) models to such data are limited. This study converts numerical data into images based on the feature importance to use the robust representation of CNN models in early diabetes diagnosis. Three different classification strategies are then applied to the resulting diabetes image data. In the first, diabetes images are fed into the ResNet18 and ResNet50 CNN models. In the second, deep features of the ResNet models are fused and classified with support vector machines (SVM). In the last approach, the selected fusion features are classified by SVM. The results demonstrate the robustness of diabetes images in the early diagnosis of diabetes.

A New Preprocessing Method for Diabetes and Biomedical Data Classification

People of all ages and socioeconomic levels, all over the world, are being diagnosed with type 2 diabetes at rates that are higher than they have ever been. It is possible for it to be the root cause of a wide variety of diseases, the most notable of which include blindness, renal illness, kidney disease, and heart disease. Therefore, it is of the utmost importance that a system is devised that, based on medical information, is capable of reliably detecting patients who have diabetes. We present a method for the identification of diabetes that involves the training of the features of a deep neural network between five and 10 times using the cross-validation training mode. The Pima Indian Diabetes (PID) data set was retrieved from the database that is part of the machine learning repository at UCI. In addition, the results of ten-fold cross-validation show an accuracy of 97.8%, a recall OF 97.8%, and a precision of 97.8% for PIMA dataset using RF algorithm. This research examined a variety of other biomedical datasets to demonstrate that machine learning may be used to develop an efficient system that can accurately predict diabetes. Several different types of machine learning classifiers, such as KNN, J48, RF, and DT, were utilized in the experimental findings of biological datasets. The findings that were obtained demonstrated that our trainable model is capable of correctly classifying biomedical data. This was demonstrated by achieving higher 99% accuracy, recall, and precision for parikson dataset.

Improving SVM performance for type II diabetes prediction with an improved non-linear kernel: Insights from the PIMA dataset

Type 2 diabetes is a chronic metabolic disease that affects a significant portion of the worldwide people. Prediction of this disease using different machine learning (ML) based algorithms has gained substantial attention due to its potential for early detection and effective intervention. One of the most powerful ML algorithm support vector machines (SVM) has proven to be effective in a variety of classification tasks, including diabetes prediction. However, the kernel function chosen has a substantial effect on the performance of SVM classifiers. This paper proposes an improved non-linear kernel for the SVM model to enhance Type 2 diabetes classification. The new kernel uses radial basis function (RBF) and RBF city block kernels that enable SVM to learn complex decision boundaries and adapt to the intricacies of the PIMA dataset. The PIMA dataset contains various clinical and demographic features of individuals. To address missing values and outliers, we impute them using the median, ensuring the integrity of the dataset. We tackle the class imbalance issue by leveraging a robust synthetic-based over-sampling approach.A comparative analysis is performed against several existing kernel functions to show that the proposed approach is superior in terms of various prediction evaluation matrices. Our recommended integrated kernel model also showed improved performance (ACC = 85.5, Recall = 87.0, Precision = 83.4, F1 score = 85.2, and AUC = 85.5) when compared to other approaches in the literature. The results of this study indicate that the proposed non-linear kernel in SVM outperforms existing kernel functions for predicting Type 2 diabetes using the PIMA dataset. Furthermore, a simulation study is carried out to robustify the proposed kernel in SVM and perform well. The improved accuracy and robustness of the model suggest its potential utility in clinical settings, aiding in the early identification and management of individuals at risk for developing diabetes.

Predictive Modeling and Analytics for Diabetes using Hyperparameter tuned Machine Learning Techniques

Accuracy of a classifier is important for the success of any prediction model. The more accuracy a classifier possesses, the more robust the system is made on it. In this paper, a disease prediction model is developed in Python for the classification of diabetes in patients. In the research paper, study is performed to make a comparative analysis of the performance of machine learning classification algorithms. The classifier's performances are enhanced by of tuning the hyperparameters of classifiers and applied different dataset preprocessing methods. In this experimental analysis, four models have been created, and each model is based on a dataset, obtained by different preprocessing methods of PIMA dataset. For each model, K-Nearest Neighbors, Decision Tree, Random Forest, and Support vector machines classification algorithms, have been applied and classifier's hyperparameters are tuned to get better results from these models.A detail analysis has also performed to get the best prediction model, the best classifier and effective preprocessing methods for it. The prediction model use F1score as the main metric. The highest F1score and accuracy are 75.68 % and 88.61% respectively, which is achieved by Random Forest classifier for dataset model D3 obtained by removing the samples having missing or unknown values from PIMA dataset.

A Diabetes Monitoring System and Health-Medical Service Composition Model in Cloud Environment

Diabetes is a common chronic illness or absence of sugar in the blood. The early detection of this disease decreases the serious risk factor. Nowadays, Machine Learning based cloud environment acts as a vital role in disease detection. The people who belong to the rural areas are not getting the proper health care treatments. So, this research work proposed an automated eHealth cloud system for detecting diabetes in the earlier stage to decrease the mortality rate and provides health treatment facilities to rural peoples. Extreme Learning Machine (ELM) is a type of Artificial Neural Network (ANN) that has a lot of potential for solving classification challenges. This research work is consisting of several activities like feature normalization, feature selection and classification. We have employed principal component analysis (PCA) for feature selection and extreme learning machine (ELM) for classification. Finally, a cloud computing-based environment with three numbers of virtual machines (vCPU-4, vCPU-8, and vCPU-16), is used for the detection of diabetes. The efficacy of the proposed model has been evaluated with the PIMA dataset in both standalone and cloud environments and achieved 90.57 % accuracy, 82.24 % sensitivity, 73.23 % specificity, and 75.03 % F-1 score with the virtual machine vCPU-16. The experimental results define the proposed model as superior to other state-of-art models with better classification accuracy and less number of features.

Hard Voting Meta Classifier for Disease Diagnosis using Mean Decrease in Impurity for Tree Models

To predict and detect various diseases, machine learning techniques are increasingly being used in the field of medical science. This study puts forward a bagging meta-estimator and feed forward neural network based voting ensemble with mean decrease in impurity feature selection to classify the disease datasets. The work was carried out using the Jupyter notebook data analysis tool, and Python 3 is used as a programming language. In this study, two chronic disease datasets - Indian Liver Patient dataset and the PIMA Indians diabetes dataset are used for building and testing the proposed model. The datasets are split into training and testing data in the ratio of 70:30. The experimental results illustrate that our proposed voting ensemble has an improved performance compared to the individual base learners. We also compared the accuracy of the model before and after the application of feature reduction technique. The results revealed that the accuracy increased with the removal of unimportant features. By using the proposed ensemble model, the average MSE, bias and variance were calculated as 0.311, 0.217 and 0.094 respectively for ILPD dataset. Similarly for PIMA dataset, the average MSE, bias and variance were calculated as 0.233, 0.186 and 0.047 respectively. These statistical parameters record a low score for ensemble classifier as compared to the individual constituent classifiers.

A modified mayfly-SVM approach for early detection of type 2 diabetes mellitus

<p>Diabetes mellitus is a chronic disease that affects many people in the world badly. Early diagnosis of this disease is of paramount importance as physicians and patients can work towards prevention and mitigation of future complications. Hence, there is a necessity to develop a system that diagnoses type 2 diabetes mellitus (T2DM) at an early stage. Recently, large number of studies have emerged with prediction models to diagnose T2DM. Most importantly, published literature lacks the availability of multi-class studies. Therefore, the primary objective of the study is development of multi-class predictive model by taking advantage of routinely available clinical data in diagnosing T2DM using machine learning algorithms. In this work, modified mayfly-support vector machine is implemented to notice the prediabetic stage accurately. To assess the effectiveness of proposed model, a comparative study was undertaken and was contrasted with T2DM prediction models developed by other researchers from last five years. Proposed model was validated over data collected from local hospitals and the benchmark PIMA dataset available on UCI repository. The study reveals that modified Mayfly-SVM has a considerable edge over metaheuristic optimization algorithms in local as well as global searching capabilities and has attained maximum test accuracy of 94.5% over PIMA.</p>

On The Analysis of Some Machine Learning Algorithms for the Prediction of Diabetes

Diabetes or Diabetes Mellitus (DM) is noxious diseases in the world. Diabetes is caused by obesity or high blood glucose level, lack of exercise and so forth. It can be manage if it’s detected at early state. Machine learning is the construction of computer system or program that can adapt and learn from their experience. PIMA dataset is used in this research works. The dataset contains some 9 attributes of 768 patients. There are different kinds of machine learning algorithms but in this research works we choose three algorithms which are under supervised learning. The algorithms are Logistic regression, Decision tree and Random forest. Each of these algorithms model were trained and tested. We later use some measure to compare and analyze the performance of the machine learning algorithms. The performance measures used are Accuracy, F-measure, Recall and Precision. Logistic Regression has the highest accuracy score which is 77%, also have the highest precision score 0.77 and have the highest f-measure 0.64. Decision Tree has the highest recall score 0.58

Assessing Machine Learning Methods for Predicting Diabetes among Pregnant Women

Machine Learning has a big impact on a lot of different scientific and technical disciplines, including medical research and Biophysics. Diabetes is a chronic condition marked by abnormally high glucose levels and the body's inefficient utilization of insulin. Diabetes is now becoming a leading cause of death all over the world. The objective of this article was to use multiple. Machine Learning methods are used to create a model with a limited number of dependencies, which could be used to study diabetic patients and diagnose diabetes using the PIMA dataset. Some of the most well-known prediction algorithms employed in this system are SVM (support vector machine), Multinomial Naive Bayes, Random forest, and Decision tree. Use these algorithms to construct a gathering of models by combining multiple combinations into one. This will enhance performance and accuracy.