With the advent of modern technologies, the healthcare industry is moving towards a more personalised smart care model. The enablers of such care models are the Internet of Things (IoT) and Artificial Intelligence (AI). These technologies collect and analyse data from persons in care to alert relevant parties if any anomaly is detected in a patient’s regular pattern. However, such reliance on IoT devices to capture continuous data extends the attack surfaces and demands high-security measures. Both patients and devices need to be authenticated to mitigate a large number of attack vectors. The biometric authentication method has been seen as a promising technique in these scenarios. To this end, this paper proposes an AI-based multimodal biometric authentication model for single and group-based users’ device-level authentication that increases protection against the traditional single modal approach. To test the efficacy of the proposed model, a series of AI models are trained and tested using physiological biometric features such as ECG (Electrocardiogram) and PPG (Photoplethysmography) signals from five public datasets available in Physionet and Mendeley data repositories. The multimodal fusion authentication model shows promising results with 99.8% accuracy and an Equal Error Rate (EER) of 0.16.

The popularity of Internet of Things (IoT) devices in smart homes has raised significant concerns regarding data security and privacy. Traditional machine learning (ML) methods for anomaly detection often require sharing sensitive IoT data with a central server, posing security and efficiency challenges. In response, this paper introduces FedGroup, a novel Federated Learning (FL) method inspired by FedAvg. FedGroup revolutionizes the central model’s learning process by updating it based on the learning patterns of distinct groups of IoT devices. Our experimental results demonstrate that FedGroup consistently achieves comparable or superior accuracy in anomaly detection when compared to both federated and non-federated learning methods. Additionally, Ensemble Learning (EL) collects intelligence from numerous contributing models, leading to enhanced prediction performance. Furthermore, FedGroup significantly improves the detection of attack types and their details, contributing to a more robust security framework for smart homes. Our approach demonstrates exceptional performance, achieving an accuracy rate of 99.64% with a minimal false positive rate (FPR) of 0.02% in attack type detection, and an impressive 99.89% accuracy in attack type detail detection.

The exposure of business applications to the web has considerably increased the variability of its workload patterns and volumes as the number of users/customers often grows and shrinks at various rates and times. Such application characteristics have increasingly demanded the need for flexible yet inexpensive computing infrastructure to accommodate variable workloads. The on-demand and per-use cloud computing model, specifically that of public Cloud Infrastructure Service Offerings (CISOs), has quickly evolved and adopted by majority of hardware and software computing companies with the promise of provisioning utility-like computing resources at massive economies of scale. However, deploying business applications on public cloud infrastructure does not lead to achieving desired economics and elasticity gains, and some challenges block the way for realizing its real benefits. These challenges are due to multiple differences between CISOs and application’s requirements and characteristics. This article introduces a detailed analysis and discussion of the economics and elasticity challenges of business applications to be deployed and operate on public cloud infrastructure. This includes analysis of various aspects of public CISOs, modeling and measuring CISOs’ economics and elasticity, application workload patterns and its impact on achieving elasticity and economics, economics-driven elasticity decisions and policies, and SLA-driven monitoring and elasticity of cloud-based business applications. The analysis and discussion are supported with motivating scenarios for cloud-based business applications. The paper provides a multi-lenses overview that can help cloud consumers and potential business application’s owners to understand, analyze, and evaluate important economics and elasticity capabilities of different CISOs and its suitability for meeting their business application’s requirements.

Recommender systems have been successfully used in many domains with the help of machine learning algorithms. However, such applications tend to use multi-dimensional user data, which has raised widespread concerns about the breach of users’ privacy. Meanwhile, wearable technologies have enabled users to collect fitness-related data through embedded sensors to monitor their conditions or achieve personalized fitness goals. In this article, we propose a novel privacy-aware personalized fitness recommender system. We introduce a multi-level deep learning framework that learns important features from a large-scale real fitness dataset that is collected from wearable Internet of Things (IoT) devices to derive intelligent fitness recommendations. Unlike most existing approaches, our approach achieves personalization by inferring the fitness characteristics of users from sensory data, minimizing the need for explicitly collecting user identity or biometric information, such as name, age, height, and weight. Our proposed models and algorithms predict (a) personalized exercise distance recommendations to help users to achieve target calories, (b) personalized speed sequence recommendations to adjust exercise speed given the nature of the exercise and the chosen route, and (c) personalized heart rate sequence to guide the user of the potential health status for future exercises. Our experimental evaluation on a real-world Fitbit dataset demonstrated high accuracy in predicting exercise distance, speed sequence, and heart rate sequence compared with similar studies. ¹ Furthermore, our approach is novel compared with existing studies, as it does not require collecting and using users’ sensitive information. Thus, it preserves the users’ privacy.

Stock prices forecasting has always been a challenging task. Although many research projects adopt machine learning and deep learning algorithms to address the problem, few of them pay attention to the varying degrees of dependencies between stock prices. In this paper we introduce a hybrid model that improves stock price prediction by emphasizing the dependencies between adjacent stock prices. The proposed model, ResNLS, is mainly composed of two neural architectures, ResNet and LSTM. ResNet serves as a feature extractor to identify dependencies between stock prices across time windows, while LSTM analyses the initial time‐series data with the combination of dependencies which considered as residuals. In predicting the SSE Composite Index, our experiment reveals that when the closing price data for the previous five consecutive trading days is used as the input, the performance of the model (ResNLS‐5) is optimal compared to those with other inputs. Furthermore, ResNLS‐5 outperforms vanilla CNN, RNN, LSTM, and BiLSTM models in terms of prediction accuracy. It also demonstrates at least a 20% improvement over the current state‐of‐the‐art baselines. To verify whether ResNLS‐5 can help clients effectively avoid risks and earn profits in the stock market, we construct a quantitative trading framework for back testing. The experimental results show that the trading strategy based on predictions from ResNLS‐5 can successfully mitigate losses during declining stock prices and generate profits in the periods of rising stock prices.

Multi-way data analysis has become an essential tool for capturing underlying structures in higher-order data sets where standard two-way analysis techniques often fail to discover the hidden correlations between variables in multi-way data. We propose a multi-objective variational autoencoder (MO-VAE) method for smart infrastructure damage detection and diagnosis in multi-way sensing data based on the reconstruction probability of autoencoder deep neural network (ADNN). Our method fuses data from multiple sensors in one ADNN at which informative features are being extracted and utilized for damage identification. It generates probabilistic anomaly scores to detect damage, asses its severity and further localize it via a new localization layer introduced in the ADNN. We evaluated our method on multi-way laboratory-based and real-life structural datasets in the area of structural health monitoring for damage diagnosis purposes. The data was collected from our deployed data acquisition system on a cable-stayed bridge in Western Sydney, a reinforced concrete cantilever beam which replicates one of the major structural components on the Sydney Harbour Bridge and a laboratory based building structure obtained from Los Alamos National Laboratory (LANL). Experimental results show that the proposed method can accurately detect structural damage. It was also able to estimate the different levels of damage severity, and capture damage locations in an unsupervised aspect. Compared to the state-of-the-art approaches, our proposed method shows better performance in terms of damage detection and localization.