Artificial Intelligence (AI) can now make very real images and videos. This helps create digital content, but it also brings big security risks. People can make fake news easily. The people need ways to find these fakes. Passive detection is a good method. It does not need watermarks added before making the image. Instead, it looks directly at the media files to find mistakes made by the computer. This paper reviews different passive detection methods. For images, this paper looks at pixel patterns and frequency data. For videos, this paper checks if frames connect smoothly over time and looks for body signals like heartbeats. Right now, detection programs work well on things they have seen before. However, they usually fail on new types of AI fakes. Future work must fix this problem so detectors can find any fake media, no matter how it was made. This paper aims to classify and review existing passive detection methods, reveal the common shortcomings of current algorithms in generalization ability, and point out the necessary path to build a general forgery detector in the future to address the security challenges brought about by the continuous evolution of deepfake technology.

The low-altitude economy, as an emerging industry form covering multiple fields such as aviation manufacturing, logistics transportation, and urban services, has become an important engine driving regional economic upgrading and industrial structure optimization. Accurately identifying low-altitude concept listed companies is an important foundation for conducting industry trend analysis, enterprise value assessment, and precise policy support. The machine learning classification algorithm, with its outstanding data mining and pattern recognition capabilities, has been widely applied in areas such as enterprise attribute prediction and industry classification, providing strong technical support for the identification of low-altitude concept listed companies. In response to the problem of low classification accuracy of low-altitude concept listed companies under high-dimensional imbalanced data, this paper proposes the KPCA-ISSA-RF classification algorithm. First, it conducts correlation analysis and violin plot analysis, and then uses multiple machine learning algorithms for comparative research. Experimental results show that the KPCA-ISSA-RF algorithm proposed in this paper has significantly better comprehensive performance than the AdaBoost, GBDT, decision tree, CatBoost, random forest, ExtraTrees, XGBoost, KNN, and logistic regression algorithms. Its accuracy and recall rate reach 92.9%, and the precision and F1 value are both 91.6%, ranking first among all algorithms. It demonstrates strong classification discrimination ability and comprehensive fitting effect, providing reliable technical support for related research and practice in the low-altitude economy.

In the wave of Industry 4.0, industrial Internet of Things (iot) devices are increasingly widely used in manufacturing, energy, chemical and other fields. The scale and operational complexity of these devices are constantly rising, and their stable operation is crucial for production efficiency and safety. To make up for the shortcomings of existing algorithms in long time series feature extraction and complex association mining, this paper proposes the Transformer-BiGRU classification and regression algorithm, and first conducts correlation analysis and violin plot analysis. Experiments show that the core evaluation indicators of the classification algorithm have significant advantages. The accuracy rate, recall rate, and precision rate all reach 84%, 84%, and 86% respectively. The F1 value is 83%, all higher than all comparison machine learning algorithms. The AUC reaches 93%, although slightly lower than CatBoost's 94%, it is higher than other algorithms such as Random Forest. Strong generalization and category discrimination capabilities; The MSE of the regression algorithm is 16.598, the RMSE is 4.074, and the MAE is 2.615, all of which are the lowest. The MAPE is 51.468, which is in a relatively low range. The R² reaches 0.442, which is significantly better than the traditional algorithm. This algorithm integrates the global dependency capture capability of Transformer with the temporal feature extraction advantages of BiGRU, providing a reliable solution for the precise analysis and prediction of the operating status of industrial Internet of Things devices, which is of great significance for ensuring the efficiency and safety of industrial production.

This study aims to improve the early prediction of academic paper impact. To this end, a dual-pathway dynamic heterogeneous graph neural network model is proposed. The model constructs a temporal heterogeneous graph consisting of multiple types of nodes, including papers, authors, institutions, and journals. Citation and non-citation pathways are designed to model knowledge diffusion relations and social-semantic associations, respectively. In addition, a life-cycle-aware mechanism is introduced to capture feature variations across the emerging, growing, and mature stages of papers. On this basis, the prediction outcome is further disentangled into three independent components, namely diffusion effect, social bias effect, and intrinsic value effect, thereby enhancing the interpretability of the model. Experimental results on datasets from multiple disciplines, including computer science, chemistry, and psychology, demonstrate that the proposed method outperforms existing mainstream models in terms of mean absolute log error and log-transformed coefficient of determination. It is especially more accurate and stable in cold-start scenarios. The results indicate that the proposed method can provide effective support for the early identification of high-potential papers.