One of the most common diseases in the world is the chronic diabetes. Diabetes has a direct impact on the lives of millions of people worldwide. Diabetes can be controlled and improved with early diagnosis, but the majority of patients continue to live with it. There is a dispirit need to a system to anticipate and select the people who are most likely to be diabetes in the future. Diagnosing the future diseased person without taking any blood or glucose screening tests, is the main goal of this study. This paper proposed a deep-learning model for diabetes disease prediction. The proposed model consists of three main phases, data pre-processing, feature selection and finally different classifiers. Initially, during the data pre-processing stage, missing values are handled, and data normalization is applied to the data. Then, three techniques are used to select the most important features which are mutual information, chi-squared and Pearson correlation. After that, multiple machine learning classifiers are used. Four experiments are then conducted to test our models. Additionally, the effectiveness of the proposed model is evaluated against that of other well-known machine learning techniques. The accuracy, AUC, sensitivity, and F-measure of the linear regression classifier are higher than those of the other methods, according to experimental data, which show that it performs better. The suggested model worked better than traditional methods and had a high accuracy rate for predicting diabetic disease.

Recently data clustering algorithm under machine learning are used in ‘real-life data’ to segregate them based on the outcome of a phenomenon. In this paper, diabetes is detected from pathological data of 768 patients using four clustering algorithms: Fuzzy C-Means (FCM), K-means clustering, Fuzzy Inference system (FIS) and Support Vector Machine (SVM). Our main objective is to make binary classification on the data table in a sense that presence or absence of diabetes of a patient. We combined the four machine learning algorithms based on entropy-based probability to enhance accuracy of detection. Before applying combining scheme, we reduce the size of variables applying multiple linear regression (MLR) on the table then logistic regression is again applied on the resultant data to keep the outlier within a narrow range. Finally, entropy based combining scheme with some modification is applied on the four ML algorithms and we got the accuracy of detection about 94% from the combining technique.

Research efforts in the prediction of stroke and diabetes prioritize early detection in order to enhance patient outcomes. To achieve this, a variety of methodologies are integrated. Existing studies, on the other hand, are marred by imbalanced datasets, lack of diversity in their datasets, potential bias, and inadequate model comparisons; these flaws underscore the necessity for more comprehensive and inclusive research methodologies. This paper provides a thorough assessment of machine learning algorithms in the context of early detection and diagnosis of stroke and diabetes. The research employed widely used algorithms, including Logistic Regression, Decision Tree, Random Forest, K-Nearest Neighbors (KNN), and XGBoost Classifier, to examine medical data and derive significant findings. The XGBoost Classifier demonstrated superior performance, with an outstanding accuracy, precision, recall, and F1-score of 87.5%. The comparative examination of the algorithms indicated that the Decision Tree, Random Forest, and XGBoost classifiers consistently exhibited strong performance across all measures. The models demonstrated impressive discrimination capabilities, with the XGBoost Classifier and Random Forest reaching accuracy rates of roughly 87.5% and 86.5% respectively. The Decision Tree Classifier exhibited notable performance, with an accuracy rate of 83%. The overall accuracy of the models was evident in the F1-score, a metric that incorporates recall and precision, where the XGBoost model exhibited a marginal improvement of 2% over the Random Forest and Decision Tree models, and 4.25 percent over the last two. The aforementioned results underscore the effectiveness of the XGBoost Classifier, which will be employed as a predictive model in this study, alongside the Random Forest and Decision Tree models, for the accurate identification of stroke and diabetes. Furthermore, combining datasets improves model performance by utilizing relative features. This integrated dataset improves the model's efficiency and creates a resilient and comprehensive prediction model, improving healthcare outcomes. The findings of this research make a valuable contribution to the advancement of AI-driven diagnostic systems, hence enhancing the quality of healthcare decision-making.

A secured smart office system is the one that is capable of recognizing and granting access to authorized persons only and manage the office appliances autonomously. The goals are access control, security and automation. Over the years, several studies have been carried out to meet these needs using RFID cards, access codes and biometrics resulting in weak security with long computational period. Switching of electrical appliances and smoke detection in case of fire outbreak were used but real time electrical appliances management that could prevent fire outbreak is yet to be achieved. This research focused its attention on the design and implementation of a smart office system that meet these needs. The system was developed using a raspberry pi 4 board. Ultrasonic sensor, camera, servo motor, relay, current and voltage sensors were interfaced with the raspberry pi for image capturing, opening the door, switching and power monitoring respectively. The system captures the image of an approaching person and process it for recognition using FaceNet; an open source model for face recognition. Information was transmitted via SIM800L GSM module as SMS to the administrator. The system shuts down the office electrical network once the supply voltage exceeds 220v ac or less than 161v ac, thus preventing any chance of fire outbreak due to irregular power supply. The accuracy of image recognition model was 93.13%. This research has shown a simple way of implementing an autonomous smart office system that is capable of providing adequate security, efficiency and convenience in offices.

Alzheimer's Disease (AD) is a progressive neurodegenerative disorder affecting millions worldwide. Early and accurate AD detection is crucial for timely intervention and improving patient outcomes. Lately, there have been notable advancements in using deep learning approaches to classify neuroimaging data associated with Alzheimer's disease. These methods have shown substantial progress in achieving accurate classification results. Nevertheless, the concept of end-to-end learning, which has the potential to harness the benefits of deep learning fully, has yet to garner extensive focus in the realm of neuroimaging. This is attributed mainly to the persistent challenge in neuroimaging, namely the limited data availability. This study employs neuroimages and Transfer Learning (TL) to identify early signs of AD and different phases of cognitive impairment. By employing transfer learning, the study uses Magnetic Resonance Imaging (MRI) images from the Alzheimer's Disease Neuroimaging (ADNI) database to classify images into various categories, such as Cognitively Normal (CN), Early Mild Cognitive Impairment (EMCI), Mild Cognitive Impairment (MCI), Late Mild Cognitive Impairment (LMCI), and Alzheimer's Disease (AD). The classification task involves training and testing three pre-trained networks: VGG-19, ResNet-50, and Inception V3. The study evaluates the performance of these networks using the confusion matrix and its associated metrics. Among the three models, ResNet-50 achieves the highest recall rate of 99.25%, making it more efficient in detecting the early stages of AD development. The study further examines the performance of the pre-trained networks on a class-by-class basis using the parameters derived from the confusion matrix. This comprehensive analysis provides insights into how each model performs for different classes within the AD classification framework. Overall, the research underscores the potential of deep learning and transfer learning in advancing early AD detection and emphasizes the significance of utilizing pre-trained models for this purpose.