This paper investigates the performance of widely used pre-trained CNN architectures (VGG16, MobileNetV3, DenseNet121, and RegNet040) across diverse datasets, particularly focusing on tuberculosis (TB) detection using Chest X-Rays (CXRs). Deep learning (DL) techniques applied to CXRs aid radiologists in promptly and accurately identifying TB, which is especially critical in low-income regions with constrained diagnostic resources. The research reveals that MobileNetV3 consistently demonstrates superior performance compared to other architectures.
Forest fires are a prevalent hazard in forests that significantly damage wildlife and the environment. It may be averted if a comprehensive system is installed in forest regions to detect fires and inform firefighting authorities to take timely action. The goal of this project is to create an Internet of Things (IoT)-based real-time detection system that detects fires and sends emergency notices to authorities. A GSM/GPRS module interacts with an IoT server because network bandwidth is typically very poor or nonexistent in forest regions. As a result, a 2G network is ideal for communicating with the server. A real-time fire monitoring system that differentiates fire and smoke is used to identify an actual fire occurrence. The nano-based Atmega328 IoT gateway identifies the forest’s fire as soon as possible and acts rapidly before it spreads across a large area. Here, this uses machine learning algorithms to detect the event of a fire in a large region. The system uses a flame sensor to detect the flame and a temperature sensor to measure the temperature in a particular area. By using the machine learning method, the system can easily give the result with actual detection and intimate the authority about the temperature condition. The data are then sent to the cloud-based application. In the event of an unusual rise in forest temperature, this will alert the forest authorities and sound the fire alarm. It can also predict future fires by using machine learning. This is accomplished using the fog computing method. Due of the sensors’ efficiency, this might potentially be applied in industrial settings. Any type of forest can make use of it. From this experiment, this research study deduced that it has a remarkable accuracy of 98% in predicting forest fires. As a result, the possibility of a false alarm is significantly decreased.
In this paper, we gave a concise note on vague fuzzy sets. We present two applications on vague sets namely an application of vague fuzzy sets in career determination using an assumed data. The application was conducted with the aid of a new distance measure of vague fuzzy sets. Also the second one deals with research questionnaire construction, filling, analysis, and interpretation is given. Respondent's decision is obtained assuming questionnaire is distributed among respondents. The respondent's decision is converted into vague data set, analysed, and from which interpretation is drawn.
Mobile Wireless Sensor Networks (MWSNs) comprise mobile sensor nodes that energetically exchange information between themselves. MWSN is self-configuring because of its dynamic nature, and each node functions with inadequate energy. Thus, energy depletion and link weight are significant challenges since links are unreliable. Uneven link strength initiates through the mobility in MWSN reasons network function. To solve this concern, this introduces Relay Awake Feature-based Efficient Route Formation (RAFE) in MWSN. RAFE approach uses a communication key function that isolates the malicious nodes in the MWSN. This measures the link weight through the sensor node packet obtained rate, loss rate, and delay factors. RAFE chooses the relay sensor node by the highest remaining energy and the greatest link weight. This approach reduces unwanted energy utilization and improves the network lifetime. Simulation results demonstrate that this approach improves the network energy efficiency and minimizes the network delay. In addition, it improves the network throughput in the network.
A numerical study is executed to analyze the steady-state heatlines visualization, fluid flow, and heat transfer inside a square enclosure with the presence of the magnetic field. The enclosure is divided into three layers, the right and left layers are filled with (Cu-Water) nanofluid while the center layer is sinusoidal porous and filled with the same nanofluid. Constant hot and cold temperature is applied to the right and left walls, respectively, the top and bottom walls are adiabatic. Galerkin finite element approach based on weak formulation is applied to solve the governing equations. The parameters studied are the number of undulation (N=1, 2 and 3), Rayleigh number (10(3)<= Ra <= 10(6)), Darcy number (10(-5)<= Da <= 10(-1)), Hartmann number (0 <= Ha <= 100) and volume fraction (0 <=phi <= 0.06). Three cases were provided depending on the number of undulations of the porous medium layer. The results obtained that the absolute value of the maximum stream function decreases with the increase of the Hartmann number and the decrease of the Darcy number for all three cases of the wavy porous layer. Heatlines and isothermal lines increase as the Darcy number is increased. The average Nusselt number grows by increasing the Rayleigh number and decreasing the Hartmann number. The enhancement of heat transfer occurred for case (2) as the Darcy number increased at a constant Ra=10(5), Ha=40. Also, It can be concluded that there was an excellent agreement between this study and those of Hamida and Charrada, by an approximately maximum absolute error of 2.062%.
