Federated learning (FL) refers to a system of training and stabilizing local machine learning models at the global level by aggregating the learning gradients of the models. It reduces the concern of sharing the private data of participating entities for statistical analysis to be carried out at the server. It allows participating entities called clients or users to infer useful information from their raw data. As a consequence, the need to share their confidential information with any other entity or the central entity called server is eliminated. FL can be clearly interpreted as a privacy-preserving version of traditional machine learning and deep learning algorithms. However, despite this being an efficient distributed training scheme, the client’s sensitive information can still be exposed to various security threats from the shared parameters. Since data has always been a major priority for any user or organization, this article is primarily concerned with discussing the significant problems and issues relevant to the preservation of data privacy and the viability and feasibility of several proposed solutions in the FL context. In this work, we conduct a detailed study on FL, the categorization of FL, the challenges of FL, and various attacks that can be executed to disclose the users’ sensitive data used during learning. In this survey, we review and compare different privacy solutions for FL to prevent data leakage and discuss secret sharing (SS)-based security solutions for FL proposed by various researchers in concise form. We also briefly discuss quantum federated learning (QFL) and privacy-preservation techniques in QFL. In addition to these, a comparison and contrast of several survey works on FL is included in this work. We highlight the major applications based on FL. We discuss certain future directions pertaining to the open issues in the field of FL and finally conclude our work.
Cardiovascular diseases (CVDs) continue to be a prominent cause of global mortality, necessitating the development of effective risk prediction models to combat the rise in heart disease (HD) mortality rates. This work presents a novel dual-stage stacked machine learning (ML) based computational risk prediction model for cardiac disorders. Leveraging a dataset that includes eleven significant characteristics from 1190 patients from five distinct sources, five ML classifiers are utilized to create the initial prediction model. To ensure robustness and generalizability, the classifiers are cross-validated ten times. The model performance is optimized by employing two hyperparameter tuning approaches: RandomizedSearchCV and GridSearchCV. These methods aim to find the optimal estimator values. The highest-performing models, specifically Random Forest, Extreme Gradient Boost, and Decision Tree undergo additional refinement using a stacking ensemble technique. The stacking model, which leverages the capabilities of the three models, attains a remarkable accuracy rate of 96%, a recall value of 0.98, and a ROC-AUC score of 0.96. Notably, the rate of false-negative results is below 1%, demonstrating a high level of accuracy and a non-overfitted model. To evaluate the model’s stability and repeatability, a comparable dataset consisting of 1000 occurrences is employed. The model consistently achieves an accuracy of 96.88% under identical experimental settings. This highlights the strength and dependability of the suggested computer model for predicting the risk of cardiac illnesses. The outcomes indicate that employing this two-step stacking ML method shows potential for prompt and precise diagnosis, hence aiding the worldwide endeavor to decrease fatalities caused by heart disease.
Image encryption is a suitable method to protect image data. Image and text data has their unique features. The available encryption algorithms are good for text data. They may not be suitable for multimedia data. In fact the pixels of natural images are highly correlated to their neighboring pixels. Due to this strong correlation any pixel can be practically predicted from the values of its neighbors. In this article, we propose a new location transformation based encryption technique. We redistribute the pixel values to different location using affine transform technique with four 8-bit keys. The transformed image then divided into 2 pixels × 2 pixels blocks and each block is encrypted using XOR operation by four 8-bit keys. The total key size used in our algorithm is 64 bit which proves to be strong enough. The experimental results proved that after the affine transform the correlation between pixel values was significantly decreased.
13 January 2023 in Journal of Ambient Intelligence and Humanized Computing
Coronary heart disease (CHD) is the world’s leading cause of death, contributing to a high mortality rate. This emphasizes the requirement for an advanced decision support system in order to evaluate the risk of CHD. This study presents an Artificial Neural Network (ANN) based intelligent healthcare system to predict the risk of CHD. The proposed ANN model is trained using the Framingham Heart Study (FHS) dataset, which comprises 4240 data instances and 15 potential risk factors. To combat overfitting, the proposed model uses four hidden dense layers with dropout rates ranging from 0.5 to 0.2. Also, two activation functions, ReLU and LeakyReLU, are used in conjunction with four optimizers: Adam, SGD, RMSProp, and AdaDelta to fine-tune the parameters and minimize the loss functions. Moreover, three sophisticated preprocessing methods, SMOTE, SMOTETomek, and SMOTEENN, along with the proposed two-stage sampling approach, are applied to address the target class data imbalance. Experimental results demonstrate that the Adam optimizer coupled with the ReLU activation function and the combined sequential effect of SMOTEENN and SMOTETomek’s two-stage sampling approach achieved superior performance. The validation accuracy reached 96.25% with a recall value of 0.98, outperforming existing approaches reported in the literature. The combined effect of approaches will be evidence of the modern healthcare decision-making support system for the risk prediction of CHD.