The rapid growth of big data and complex machine learning models—gradient-boosted trees and deep neural networks—has produced highly accurate but opaque"black-box"predictors across medicine, finance, and industry, making interpretability a central concern in data analysis. SHapley Additive exPlanations (SHAP), grounded in cooperative game theory, has become one of the most influential interpretability methods because it provides theoretically consistent feature attributions at both the local and global levels. This paper presents a systematic literature review of SHAP and its role in data analysis. It synthesizes SHAP's theoretical foundations, its main implementations (TreeSHAP, KernelSHAP, and DeepSHAP), its visualization toolkit, and its practical applications, and it reports a compact empirical study comparing the three explainers on a clinical dataset. This study finds that, although SHAP markedly improves transparency and decision support, open challenges remain in computational cost, the reliability of explanations under feature correlation, and consistency across methods. The significance of this work is twofold: theoretically, it organizes SHAP's variants and properties within a single coherent framework; practically, it offers data analysts a structured, evidence-based reference for selecting and applying SHAP appropriately, thereby supporting more transparent, reliable, and accountable model-driven decisions.

Parametric models widely used in estimation often violate assumptions in practical research. As flexible alternatives, machine learning methods, especially tree ensembles and deep neural networks, impose fewer prior assumptions on functional forms for parameter estimation. This paper systematically examines eight estimation methods—ordinary least squares, ridge regression, lasso, random forest, XGBoost, LightGBM, multilayer perceptron (MLP), and deep neural networks—across four simulation regimes (linear, semiparametric, nonlinear, and high-dimensional sparse) and two real-world datasets: the Home Credit Default Risk dataset (307,511 samples) and the PIMA Indians Diabetes dataset (768 samples). This study evaluate model bias, mean squared error (MSE) and computational overhead to determine the applicable scenarios for each method category. Experimental results demonstrate that tree-based methods perform steadily across various scenarios. Although deep neural networks incur higher computational overhead, they achieve the minimal MSE when facing strong nonlinearity with moderate or large sample sizes. These findings provide actionable guidance for selecting estimation methods based on data characteristics, bridging theoretical advances in machine learning and practical estimation. The results verify that no single estimator outperforms all others across all data scenarios. The optimal selection relies on the joint effects of data nonlinearity, dimensionality and sample size, which highlights the necessity of diagnosis-oriented method selection in empirical studies.

Artificial intelligence (AI) in breast imaging has progressed from retrospective algorithm testing to prospective and real-world evaluation. In full-field digital mammography (FFDM), randomized, prospective, and implementation studies show that AI-supported reading maintains or increases cancer detection while reducing radiologist workload when thresholds, arbitration rules, and audit procedures are prespecified. Evidence for digital breast tomosynthesis (DBT) is expanding, but remains weighted toward reader studies and retrospective external validation rather than population-level deployment. Mammography-derived risk models extend AI from lesion detection to longitudinal risk stratification, provided that absolute-risk calibration and subgroup monitoring are performed locally. In breast MRI, AI triage, lesion classification, ultrafast acquisition, and reconstruction methods address access and workload constraints, although safety depends on conservative operating points and interval cancer surveillance. This paper draws the conclusion with a five-layer translation framework linking technical validity, workflow fit, safety monitoring, fairness, and economic sustainability, which offers references for further research in the applications of artificial intelligence in breast cancer imaging.

The rapid growth of online shopping platforms has significantly increased the analytical value of product review data. Short reviews on Chinese e-commerce platforms, typically containing fewer than 50 characters, simultaneously address multiple dimensions, including product quality, logistics, customer service, and value for money, rendering conventional sentence-level sentiment classification insufficient for fine-grained analysis. This study adopts aspect-based sentiment analysis (ABSA) as the core framework and constructs a labeling schema comprising four sentiment dimensions and three polarity levels, yielding 12 classification categories. Rule-based automated annotation was applied to 62,770 publicly available Chinese consumer reviews; manual verification on a 500-sample subset confirmed an annotation accuracy of 87.2%. Three lightweight models—TextCNN, BiLSTM with attention (BiLSTM+Attn), and ERNIE 3.0-Nano—were trained and evaluated under identical experimental conditions on an NVIDIA RTX 4060 laptop GPU. ERNIE 3.0-Nano proved to be the most accurate, with an accuracy of 0.6363 and a macro-average F1 score of 0.5798. BiLSTM+Attn was the fastest on GPU: 0.004 ms per sample. TextCNN had the smallest parameter count: 0.69 million. These findings provide quantitative evidence for model selection under diverse deployment constraints.

With the rapid deployment of neural machine translation (NMT) and large language models (LLMs), AI-assisted translation has become a cornerstone of multilingual communication. Despite achieving impressive fluency, these systems often perpetuate subtle yet systematic cultural biases embedded in training corpora, model architectures, and inference pipelines. This paper presents a systematic review of cultural bias in AI translation, organized around three research questions: (1) how cultural bias manifests, (2) how it can be identified, and (3) how it can be mitigated. Drawing on recent advances in machine translation, multilingual NLP, and AI fairness, this study analyzes manifestations across gendered stereotyping, religious oversimplification, regional framing, and cultural normalization; and then synthesizes detection methods, including benchmark-based evaluation, contrastive probing, embedding association tests, and human-in-the-loop assessment. For mitigation, this paper proposes a five-layer framework spanning data auditing, model adaptation, inference-time intervention, post-editing, and governance. To validate the framework, we conduct five proof-of-concept experiments: cross-lingual gender bias detection with statistical testing, systematic cultural fidelity evaluation under prompt engineering, contrastive sentiment analysis under high-/low-risk contexts, word embedding association tests (WEAT) with permutation-based significance, and an integrated audit pipeline with automated mitigation. Results demonstrate significant gender bias (χ²=29.99, p<0.001), a pervasive "male-as-default" phenomenon, significant gains from culture-aware prompting (p=0.03), and robust embedding-space bias (permutation test p=0.0001). The audit pipeline successfully integrates detection and mitigation into an actionable workflow. We conclude by outlining future directions for low-resource languages, intersectional bias, and production-level deployment.

This paper examines whether layer-wise English-Chinese cross-lingual alignment in decoder-only language models is associated with controlled target-word prediction performance. We compare EleutherAI/Pythia-1.4B, Microsoft Phi-3.5-mini-instruct, and Qwen3-1.7B-Base. For each model, we extract target-word hidden representations across Transformer layers and evaluate four mapping conditions: no mapping, orthogonal Procrustes, ridge regression, and a residual multilayer perceptron. We also evaluate final-layer target-word prediction and an auxiliary multilingual language-identification task. Results show that English-Chinese alignment generally becomes stronger in deeper layers, and that ridge and MLP mappings usually outperform no mapping and Procrustes. However, alignment and prediction are related but not identical: some models show strong geometric alignment without the best target prediction. These findings suggest that representation alignment captures an important internal property of multilingual models, while final prediction also depends on tokenization, output distributions, instruction tuning, and training data.

The ability to efficiently analyze large-scale and complex biological data helps artificial intelligence to rapidly revolutionize biomedical research. This paper summarizes the value of artificial intelligence in three biomedical domains: multi-omics data integration for disease prediction, medical image analysis for clinical diagnose and drug discovery. In multi-omics research, artificial intelligence optimizes integration outcomes of heterogeneous biological datasets, deepening cognition of disease mechanisms and polishing predictive models. In medical imaging, deep learning performs excellently in various diagnosis tasks, yet the lack of interpretability brings out limitations. To address the problem, explainable artificial intelligence was generated to effectively increase model transparency and clinical validation. In drug discovery, artificial intelligence significantly accelerates main processes, including virtual screening, drug repurposing and protein structure prediction. In particular, AlphaFold is one of the most paramount breakthroughs. Despite these advancements, data heterogeneity, lack of clinical validation samples and model interpretability shortage restrain the practical usage of relevant techniques. This paper sorts out the potential and limitations of artificial intelligence in biomedical research, pointing out that future research should build more stable and interpretable models.

Recent advances in large artificial intelligence (AI) models have transformed the landscape of life sciences and healthcare by enabling more efficient knowledge extraction, biomedical data analysis, and clinical decision support. Compared with traditional task-specific AI models, large language models (LLMs), generative AI, and multimodal foundation models demonstrate superior capabilities in knowledge representation, reasoning, and cross-domain learning. This review summarizes recent developments in the application of large AI models in life sciences and healthcare. First, the technical foundations of large AI models and their adaptation to biomedical data are introduced. Subsequently, representative applications in life science research, clinical healthcare, and drug discovery are reviewed, with particular attention to biomedical literature analysis, biological sequence modeling, medical image interpretation, clinical decision support, and precision medicine. Furthermore, the opportunities and challenges associated with deploying large AI models in healthcare are discussed, including improvements in research efficiency, disease diagnosis, and personalized treatment, as well as concerns regarding hallucination, privacy protection, algorithmic bias, interpretability, and regulatory governance. Overall, large AI models are reshaping biomedical research and healthcare delivery by promoting more intelligent, data-driven, and integrated approaches. Future progress will depend on continuous advances in trustworthy AI, multimodal learning, and interdisciplinary collaboration to ensure safe, reliable and clinically applicable AI systems.

This paper primarily investigates the application of neural network-based emotion recognition methods in human-computer interaction, with a focus on convolutional neural networks in facial emotion recognition, vision transformers, and multimodal emotion recognition research approaches. This paper first analyzes the basic process and principle of ResNet-50 in facial image feature extraction, explaining that it can extract local facial features through convolution operations and residual connections, and use Softmax to complete emotion classification. Secondly, this paper also introduces Vision Transformer, which models global features through image patch partitioning and a self-attention mechanism, and analyzes its advantages in capturing the correlation between different facial regions. Subsequently, this paper further discusses the application of multimodal emotion recognition in medical scenarios, demonstrating its ability to integrate various information such as facial images, speech, text, and physiological signals. It is pointed out that different neural network models are suitable for different human-computer interaction scenarios, and a single model is difficult to adapt to complex environments. Future emotion recognition technology should further develop towards model lightweighting and multimodal fusion.

This study evaluates the predictive performance of three deep learning architectures: Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), and a hybrid Convolutional Neural Network-LSTM (CNN-LSTM), for forecasting the CBOE Volatility Index (VIX). Using daily VIX data from January 2000 to June 2024, we benchmark these models using mean absolute error (MAE), mean squared error (MSE), and the coefficient of determination (R2). The CNN-LSTM model consistently outperforms both unidirectional and bidirectional LSTMs, achieving the lowest MAE (1.758), MSE (9.033), and highest R2 (0.872). Contrary to expectations, the BiLSTM performs worst among the three, with an R2 of 0.670, and this suggests that bidirectional information flow may introduce noise rather than improve accuracy for volatility forecasting. The results also indicate that deep learning models maintain predictive stability during periods of market turbulence, with CNN-LSTM demonstrating the strongest resilience. These findings have practical implications for real-time risk monitoring systems in volatile financial markets.