Dynamic Anomaly Identification in Accounting Transactions via Multi-Head Self-Attention Networks

This study addresses the problem of dynamic anomaly detection in accounting transactions and proposes a real-time detection method based on a Transformer to tackle the challenges of hidden abnormal behaviors and high timeliness requirements in complex trading environments. The approach first models accounting transaction data by representing multi-dimensional records as time-series matrices and uses embedding layers and positional encoding to achieve low-dimensional mapping of inputs. A sequence modeling structure with multi-head self-attention is then constructed to capture global dependencies and aggregate features from multiple perspectives, thereby enhancing the ability to detect abnormal patterns. The network further integrates feed-forward layers and regularization strategies to achieve deep feature representation and accurate anomaly probability estimation. To validate the effectiveness of the method, extensive experiments were conducted on a public dataset, including comparative analysis, hyperparameter sensitivity tests, environmental sensitivity tests, and data sensitivity tests. Results show that the proposed method outperforms baseline models in AUC, F1-Score, Precision, and Recall, and maintains stable performance under different environmental conditions and data perturbations. These findings confirm the applicability and advantages of the Transformer-based framework for dynamic anomaly detection in accounting transactions and provide methodological support for intelligent financial risk control and auditing.

Deep Learning Approach for Clinical Risk Identification Using Transformer Modeling of Heterogeneous EHR Data

This study proposes a Transformer-based longitudinal modeling method to address challenges in clinical risk classification with heterogeneous Electronic Health Record (EHR) data, including irregular temporal patterns, large modality differences, and complex semantic structures. The method takes multi-source medical features as input and employs a feature embedding layer to achieve a unified representation of structured and unstructured data. A learnable temporal encoding mechanism is introduced to capture dynamic evolution under uneven sampling intervals. The core model adopts a multi-head self-attention structure to perform global dependency modeling on longitudinal sequences, enabling the aggregation of long-term trends and short-term fluctuations across different temporal scales. To enhance semantic representation, a semantic-weighted pooling module is designed to assign adaptive importance to key medical events, improving the discriminative ability of risk-related features. Finally, a linear mapping layer generates individual-level risk scores. Experimental results show that the proposed model outperforms traditional machine learning and temporal deep learning models in accuracy, recall, precision, and F1-Score, achieving stable and precise risk identification in multi-source heterogeneous EHR environments and providing an efficient and reliable framework for clinical intelligent decision-making.