As COVID-19 enters the pandemic stage, the resulting infections, deaths and economic shocks are emerging. To minimize anxiety and uncertainty about socio-economic damage caused by the COVID-19 pandemic, it is necessary to reasonably predict the economic impact of future disease trends by scientific means. Based on previous cases of epidemic (such as influenza) and economic trends, this study has established an epidemic disease spread model and economic situation prediction model. Based on this model, the author also predict the economic impact of future COVID-19 spread. The results of this study are as follows. First, the deep learning-based economic impact prediction model, which was built based on historical infectious disease data, was verified with verification data to ensure 77% accuracy in predicting inflation rates. Second, based on the economic impact prediction model of the deep learning-based infectious disease, the author presented the COVID-19 trend and future economic impact prediction results for the next 1 year. Currently, most of the published studies on COVID-19 are on the prediction of disease spread by statistical mathematical calculations. This study is expected to be used as an empirical reference to efficient and preemptive decision making by predicting the spread of diseases and economic conditions related to COVID-19 using deep learning technology and historical infectious disease data.

Electroencephalography (EEG)-based emotion recognition (ER) is a growing research area in affective computing and human–computer interaction. EEG signals capture non-invasive measurements of brain activity and provide real-time access to emotional processes at the neural level. Compared to facial expressions or speech, EEG is less susceptible to voluntary masking and therefore holds promise for robust emotion decoding. However, traditional EEG-based ER methods are often limited by low classification accuracy, small sample sizes, and reliance on discrete or bipolar affective labels. Recent advances in deep learning, multimodal fusion, and semantic-space emotion models offer promising directions. This paper reviews state-of-the-art EEG-based ER approaches, focusing on signal preprocessing, feature extraction, and classification techniques. In addition, we identify emerging trends and persistent challenges, including dataset generalizability, inter-subject variability, and the need for fine-grained emotion annotation. By synthesizing current methodologies and outlining future paths, this review aims to guide both researchers and practitioners in the development of more accurate, scalable, and cognitively grounded EEG-based emotion recognition systems.

Recent advances in time-series anomaly detection have leveraged artificial intelligence (AI) to improve detection performance. In industrial control systems (ICSs), however, acquiring training data is challenging due to operational constraints and the difficulty of system shutdowns. To address this, many countries are developing ICS simulators and testbeds to generate training data. This study uses a publicly available ICS testbed dataset as a benchmark for the discriminator in a Semi-Supervised Generative Adversarial Network (SGAN). The goal is to generate large volumes of synthetic time-series data through adversarial training between generator and discriminator networks, thereby mitigating data scarcity in ICS anomaly detection. Comparative experiments were conducted using this synthetic data to evaluate its impact on existing detection models. Using the HAI 22.04 dataset from the National Security Research Institute, this study performed feature engineering and preprocessing to identify correlations and remove irregularities. Various models, including One-Class SVM, VAE, CNN-GRU-Autoencoder, and CNN-LSTM-Autoencoder, were trained and tested on the dataset. A synthetic dataset was then generated via SGAN and validated using PCA and t-SNE. The results show that applying SGAN-generated data to time-series anomaly detection yielded significant performance improvements in F1 score. Additional validation using the SWaT dataset from the National University of Singapore confirmed similar gains. These findings indicate that synthetic data generated by SGANs can effectively enhance semi-supervised learning for anomaly detection, classification, and prediction in data-constrained environments such as medical, industrial, transportation, and environmental systems.