Search

In this study, we employ explainable machine learning (ML) models to predict daily streamflow ( Q flow ) by leveraging hydro-meteorological parameters. The predictive matrix incorporates crucial factors such as daily rainfall, temperature, relative humidity, solar radiation, wind speed, and the one-day lag value of Q flow . Notably, among these parameters, the one-day lag value of Q flow , along with rainfall, solar radiation, temperature, and relative humidity emerge as highly influential predictors. We apply various ML models, including bagging ensemble learning, boosting ensemble learning, Gaussian process regression (GPR), and automated machine learning (Auto ML). Following a rigorous evaluation, the bagging ensemble learning model stands out as the most effective with a correlation coefficient (R = 0.80) and root-mean-square error (RMSE = 218). Further, we compare the Q flow predicted using ML models with a process-based hydrological model (SWAT) that was executed using a similar set of climatic variables as the input parameters. In our case, the predictive strength of the ML model (R = 0.80; RMSE = 218) to estimate ( Q flow ) is greater than the SWAT (R = 0.82; RMSE = 281). In conclusion, by emphasizing the importance of explainable ML models and highlighting the significance of specific hydro-meteorological parameters, our study contributes to advancing the field of hydrology and water resource management. [ABSTRACT FROM AUTHOR]

We use Soil and Water Assessment Tool (SWAT) to simulate the combined effects of land use/land cover (LU/LC) and climate change on the hydrological response of the Upper Betwa River Catchment (UBRC), a semi-arid region in Central India. We execute this model for two different time periods, 1982–2000 and 2001–2018, using the LU/LC data of 1990 and 2018, respectively. We classified the Landsat satellite images of 1990 and 2018 to obtain the dominant LU/LC classes (water body, built-up, forest, agriculture, and open land) in the catchment. The water body, built-up areas, and cropland have increased by 63%, 65%, and 3%, respectively, whereas forest cover and open land decreased by 16% and 23% in the UBRC from 1990 to 2018. The observed climate data in UBRC shows an increase in the average temperature and decrease in the total rainfall during the period between 1980 to 2018. Once the model is set up, we perform the calibration and validation by using the SWAT Calibration Uncertainty Program (SWAT-CUP). We considered two time periods (1991–1994 and 2001–2007) for the calibration and (1995–1998 and 2008–2014) for the validation. For both these time periods, the calibration and validation result of our model is satisfactory. The output of our calibrated model shows a relative decrease in rainfall (12%), surface runoff (21%), and percolation (9%) in the catchment during the period between 2001–2018 as compared to 1982–2000. Finally, we simulate the surface runoff and percolation in the UBRC using the future climate change scenario. We used the bias-corrected multi-model ensemble of CMIP6 GCMs for four different climate scenarios (2023–2100) by assuming no change in the existing LU/LC. We do this for two different time slices: one from 2023–2060 and the other from 2061–2100. For all the climate scenarios, rainfall and surface runoff in the catchment are expected to decrease by 15–40% and 50–79% as compared to the baseline period of 1982–2018. Percolation in the catchment will have a mixed response. It is expected to decrease by 18% in the middle part of the catchment and increase about 25% in the remaining parts of the catchment. [ABSTRACT FROM AUTHOR]

We propose a new architecture based on a fully connected feed-forward Artificial Neural Network (ANN) model to estimate surface soil moisture from satellite images on a large alluvial fan of the Kosi River in the Himalayan Foreland. We have extracted nine different features from Sentinel-1 (dual-polarised radar backscatter), Sentinel-2 (red and near-infrared bands), and Shuttle Radar Topographic Mission (digital elevation model) satellite products by leveraging the linear data fusion and graphical indicators. We performed a feature importance analysis by using the regression ensemble tree approach and also feature sensitivity to evaluate the impact of each feature on the response variable. For training and assessing the model performance, we conducted two field campaigns on the Kosi Fan in December 11–19, 2019 and March 01–06, 2022. We used a calibrated TDR probe to measure surface soil moisture at 224 different locations distributed throughout the fan surface. We used input features to train, validate, and test the performance of the feed-forward ANN model in a 60:10:30 ratio, respectively. We compared the performance of ANN model with ten different machine learning algorithms [i.e., Generalised Regression Neural Network (GRNN), Radial Basis Network (RBN), Exact RBN (ERBN), Gaussian Process Regression (GPR), Support Vector Regression (SVR), Random Forest (RF), Boosting Ensemble Learning (Boosting EL), Recurrent Neural Network (RNN), Binary Decision Tree (BDT), and Automated Machine Learning (AutoML)]. We observed that the ANN model accurately predicts the soil moisture and outperforms all the benchmark algorithms with correlation coefficient (R = 0.80), Root Mean Square Error (RMSE = 0.040 m 3 / m 3 ), and bias = 0.004 m 3 / m 3 . Finally, for an unbiased and robust conclusion, we performed spatial distribution analysis by creating thirty different sets of training-validation-testing datasets. We observed that the performance remains consistent in all thirty scenarios. The outcomes of this study will foster new and existing applications of soil moisture. [ABSTRACT FROM AUTHOR]

Wireless communication networks have much data to sense, process, and transmit. It tends to develop a security mechanism to care for these needs for such modern-day systems. An intrusion detection system (IDS) is a solution that has recently gained the researcher's attention with the application of deep learning techniques in IDS. In this paper, we propose an IDS model that uses a deep learning algorithm, conditional generative adversarial network (CGAN), enabling unsupervised learning in the model and adding an eXtreme gradient boosting (XGBoost) classifier for faster comparison and visualization of results. The proposed method can reduce the need to deploy extra sensors to generate fake data to fool the intruder 1.2–2.6%, as the proposed system generates this fake data. The parameters were selected to give optimal results to our model without significant alterations and complications. The model learns from its dataset samples with the multiple-layer network for a refined training process. We aimed that the proposed model could improve the accuracy and thus, decrease the false detection rate and obtain good precision in the cases of both the datasets, NSL-KDD and the CICIDS2017, which can be used as a detector for cyber intrusions. The false alarm rate of the proposed model decreases by about 1.827%. [ABSTRACT FROM AUTHOR]

Momentous increase in the popularity of explainable machine learning models coupled with the dramatic increase in the use of synthetic data facilitates us to develop a cost-efficient machine learning model for fast intrusion detection and prevention at frontier areas using Wireless Sensor Networks (WSNs). The performance of any explainable machine learning model is driven by its hyperparameters. Several approaches have been developed and implemented successfully for optimising or tuning these hyperparameters for skillful predictions. However, the major drawback of these techniques, including the manual selection of the optimal hyperparameters, is that they depend highly on the problem and demand application-specific expertise. In this paper, we introduced Automated Machine Learning (AutoML) model to automatically select the machine learning model (among support vector regression, Gaussian process regression, binary decision tree, bagging ensemble learning, boosting ensemble learning, kernel regression, and linear regression model) and to automate the hyperparameters optimisation for accurate prediction of numbers of k-barriers for fast intrusion detection and prevention using Bayesian optimisation. To do so, we extracted four synthetic predictors, namely, area of the region, sensing range of the sensor, transmission range of the sensor, and the number of sensors using Monte Carlo simulation. We used 80% of the datasets to train the models and the remaining 20% for testing the performance of the trained model. We found that the Gaussian process regression performs prodigiously and outperforms all the other considered explainable machine learning models with correlation coefficient (R = 1), root mean square error (RMSE = 0.007), and bias = − 0.006. Further, we also tested the AutoML performance on a publicly available intrusion dataset, and we observed a similar performance. This study will help the researchers accurately predict the required number of k-barriers for fast intrusion detection and prevention. [ABSTRACT FROM AUTHOR]

Road traffic accident-induced traumatic brain injury is a critical health concern. Understanding the biomechanics of the head during accidental impact is important for devising impact mitigation and injury prevention strategies. In this work, we reconstruct the specific case of collision of the vehicle with pedestrian using the pendulum impactor setup. The surrogate head was used for studying biomechanical response. The surrogate head consisted of a hollow polycarbonate spherical shell filled with water. Dummy neck is also considered in the form of a spring. Head acceleration, skull strain, and intracranial pressures (ICPs) were measured. Using this setup, we were able to recreate head kinematics and kinetics that are implied during these accidents. The measured biomechanical response is consistent with the findings in the literature. The work provides novel insights into the biomechanics of road traffic accident-induced brain injury. [ABSTRACT FROM AUTHOR]

In the context of wireless sensor networks (WSNs), the utilization of artificial intelligence (AI)-based solutions and systems is on the ascent. These technologies offer significant potential for optimizing services in today's interconnected world. AI and nature-inspired algorithms have emerged as promising approaches to tackle various challenges in WSNs, including enhancing network lifespan, data aggregation, connectivity, and achieving optimal coverage of the targeted area. Coverage optimization poses a significant problem in WSNs, and numerous algorithms have been proposed to address this issue. However, as the number of sensor nodes within the sensor range increases, these algorithms often encounter difficulties in escaping local optima. Hence, exploring alternative global metaheuristic and bio-inspired algorithms that can be adapted and combined to overcome local optima and achieve global optimization in resolving wireless sensor network coverage problems is crucial. This paper provides a comprehensive review of the current state-of-the-art literature on wireless sensor networks, coverage optimization, and the application of machine learning and nature-inspired algorithms to address coverage problems in WSNs. Additionally, we present unresolved research questions and propose new avenues for future investigations. By conducting bibliometric analysis, we have identified that binary and probabilistic sensing model are widely employed, target and k-barrier coverage are the most extensively studied coverage scenarios in WSNs, and genetic algorithm and particle swarm optimization are the most commonly used nature-inspired algorithms for coverage problem analysis. This review aims to assist researchers in exploring coverage problems by harnessing the potential of nature-inspired and machine-learning algorithms. It provides valuable insights into the existing literature, identifies research gaps, and offers guidance for future studies in this field. [ABSTRACT FROM AUTHOR]

Superparamagnetic NiFeO catalyst immobilized palladium(II) has been synthesized using a microemulsion method and functionalized with Schiff-base group to form an immobilized bidentate ligand. The catalyst was characterized by multiple analytical techniques. The immobilized palladium complex was used in EtOH/water under aerobic conditions. The catalyst is applicable for various substrates and easily separable. The as-prepared catalyst was reused up to five runs with slight decrease in catalytic activity. Graphical Abstract: [Figure not available: see fulltext.] [ABSTRACT FROM AUTHOR]