Rainfall forecasting is crucial in the field of meteorology and hydrology. However, existing solutions always achieve low prediction accuracy for short-term rainfall forecasting. Numerical forecasting models perform worse in many conditions. Machine learning approaches neglect the influences of physical factors in upstream or downstream regions, which make forecasting accuracy fluctuate in different areas. To improve the overall forecasting accuracy for short-term rainfall, this paper proposes a novel solution called Dynamic Regional Combined short-term rainfall Forecasting approach (DRCF) using Multi-layer Perceptron (MLP). First, Principal Component Analysis (PCA) is used to reduce the dimension of thirteen physical factors, which serves as the input of MLP. Second, a greedy algorithm is applied to determine the structure of MLP. The surrounding sites are perceived based on the forecasting site. Finally, to solve the clutter interference which is caused by the extension of the perception range, DRCF is enhanced with several dynamic strategies. Experiments are conducted on data from 56 real-world meteorology sites in China, and we compare DRCF with atmospheric models and other machine learning approaches. The experimental results show that DRCF outperforms existing approaches in both threat score (TS) and root mean square error (RMSE).

Over the years, rampant wildfires have plagued the state of California, creating economic and environmental loss. In 2018, wildfires cost nearly 800 million dollars in economic loss and claimed more than 100 lives in California. Over 1.6 million acres of land has burned and caused large sums of environmental damage. Although, recently, researchers have introduced machine learning models and algorithms in predicting the wildfire risks, these results focused on special perspectives and were restricted to a limited number of data parameters. In this paper, we have proposed two data-driven machine learning approaches based on random forest models to predict the wildfire risk at areas near Monticello and Winters, California. This study demonstrated how the models were developed and applied with comprehensive data parameters such as powerlines, terrain, and vegetation in different perspectives that improved the spatial and temporal accuracy in predicting the risk of wildfire including fire ignition. The combined model uses the spatial and the temporal parameters as a single combined dataset to train and predict the fire risk, whereas the ensemble model was fed separate parameters that were later stacked to work as a single model. Our experiment shows that the combined model produced better results compared to the ensemble of random forest models on separate spatial data in terms of accuracy. The models were validated with Receiver Operating Characteristic (ROC) curves, learning curves, and evaluation metrics such as: accuracy, confusion matrices, and classification report. The study results showed and achieved cutting-edge accuracy of 92% in predicting the wildfire risks, including ignition by utilizing the regional spatial and temporal data along with standard data parameters in Northern California.

There is a growing demand for Green AI (Artificial Intelligence) technologies in the market and society, as it emerges as a promising technology. Green AI technologies are used to create sustainable solutions and reduce the environmental impact of AI. This paper focuses on describing the services of Green AI and the challenges associated with it at the community level. This article also highlights the accuracy levels of machine learning algorithms for various time periods. The process of choosing the appropriate input parameters for weather, locations, and complexity is outlined in this paper to examine the ML algorithms. For correcting the algorithm performance parameters, metrics like RMSE (root mean square error), MSE (mean square error), MAE (mean absolute error), and MPE (mean percentage error) are considered. Considering the performance and results of this review, the LSTM (long short-term memory) performed well in most cases. This paper concludes that highly advanced techniques have dramatically improved forecasting accuracy. Finally, some guidelines are added for further studies, needs, and challenges. However, there is still a need for more solutions to the challenges, mainly in the area of electricity storage.

Recently, compressive sensing (CS) and visual security (VS) have caught researchers attention in information security field. However, the measurement matrix is often reused in CS, which makes it vulnerable to chosen plaintext attack (CPA). In addition, when generating meaningful cipher images, the size of the carrier image is usually not less than the size of the plain image. In order to overcome these drawbacks, a new visually secure image encryption scheme using CS and reversible color transformation is proposed. The algorithm consists of two stages: compression and embedding. In the first stage, chaotic sequence is used to generate different structurally random matrices. When CS is performed, a random number is added during the process of sampling. By choosing different random numbers, different measurement matrices can be used to compress and encrypt the same image in different order. In the second stage, block pairing, color transformation and block replacement are employed to obtain a meaningful image. Different from the block replacement between two similar images, this paper first attempts to replace the block of the carrier image with a compressed noise-like image block. Thus, the carrier image can be smaller than the plain image, which saves the bandwidth of transmission. Both theoretical analysis and experimental results show that the proposed encryption scheme has good encryption performance, can effectively resist common attacks, and is suitable for meaningful image encryption.

According to the World Health Organization (WHO), wearing a face mask is one of the most effective protections from airborne infectious diseases such as COVID-19. Since the spread of COVID-19, infected countries have been enforcing strict mask regulation for indoor businesses and public spaces. While wearing a mask is a requirement, the position and type of the mask should also be considered in order to increase the effectiveness of face masks, especially at specific public locations. However, this makes it difficult for conventional facial recognition technology to identify individuals for security checks. To solve this problem, the Spartan Face Detection and Facial Recognition System with stacking ensemble deep learning algorithms is proposed to cover four major issues: Mask Detection, Mask Type Classification, Mask Position Classification and Identity Recognition. CNN, AlexNet, VGG16, and Facial Recognition Pipeline with FaceNet are the Deep Learning algorithms used to classify the features in each scenario. This system is powered by five components including training platform, server, supporting frameworks, hardware, and user interface. Complete unit tests, use cases, and results analytics are used to evaluate and monitor the performance of the system. The system provides cost-efficient face detection and facial recognition with masks solutions for enterprises and schools that can be easily applied on edge-devices.