Firuz Kamalov

Cybersecurity has become a great concern in many real-world applications involving adversaries with Machine Learning (ML) algorithms being more widely used. This concern is more challenging in Internet of Things (IoT) platforms. As IoT-enabled applications are growing at a rapid pace in every sector there are growing security related incidents as well. ML algorithms are widely deployed to perform data analysis, reasoning and decision-making over the data emanating from IoT devices. Security of this data while collection, communication and computing is a major challenge. Various attackers are trying to find weaknesses in the ML algorithms and trying to deceive these ML algorithms to learn the wrong information from the data. Countermeasures need to be developed to evaluate the security of the ML models. To develop such countermeasures, one needs to understand all possible attacks on the ML models. Data poisoning attacks are a class of adversarial attacks on ML where an adversary has the power to alter a small fraction of the training data in order to make the trained classifier to satisfy certain objectives. In order to develop an attack-resistant ML model, one needs to know all possible attacks on these models. The recent data poisoning technique such as the Fast Gradient Sign Method (FGSM) is static and provides very micro control to the attacker on creating adversarial data. In this research, we develop a more robust data poisoning technique for deep neural networks using Generative Adversarial Networks (GANs) to create a data poisoning attack. We then evaluate the performance of the proposed algorithm and also compare the results obtained by FGSM.

The analysis of heat transmission and fluid flow characteristics within the cavity is useful to improve the features of several applications including energy storage devices and hybrid fuel cells. With this motivation, the present model investigates the characteristics of magneto-convective heat transmission and fluid flow within a square porous enclosure with hot and cold slits. The heat transfer features of electrically conducting hybrid nanofluids Fe3O4–MWCNT– water and Fe3O4–MWCNT– kerosene are analyzed inside the enclosure. The non-Fourier thermal flux model is deployed, and the internal heat absorption/generation effect is considered. The marker-and-cell numerical scheme is adopted to solve the transformed dimensionless mathematical model with associated initial–boundary conditions. An exhaustive parametric investigation is implemented to estimate the influence of key parameters on transport phenomena. The computations show that augmenting the Hartmann number values modifies the fluid flow and temperature features substantially for both hybrid nanofluids. Enhancing the values of nanoparticles volume fraction promotes the heat transfer. When 5% Fe3O4–MWCNT nanoparticles are suspended into water and kerosene base fluids, Fe3O4–MWCNT– kerosene hybrid nanofluid achieves 6.85% higher mean heat transfer rate compared to Fe3O4–MWCNT– water hybrid nanoliquid. In the existence of heat absorption, the mean rate of heat transfer of Fe3O4–MWCNT– water hybrid nanofluid is 78.92% lower than Fe3O4–MWCNT– kerosene hybrid nanoliquid. Greater energy transmission is noticed in the case of Fe3O4–MWCNT– kerosene hybrid nanofluid, and the enhanced fluid flow is noticed in the case of Fe3O4–MWCNT– water hybrid nanofluid. Fourier's model (δe=0) estimates higher heat transfer rate than that of the Cattaneo–Christov (non-Fourier) heat flux model (δe≠0).