With the shift toward electrification, lightweight architectures, and intelligent systems in the automotive sector, demand for high-performance, tailor-made components is rising, while traditional manufacturing methods struggle to cope with complex geometries and accelerated design cycles. Therefore, this study investigates the application of additive manufacturing technology in the design optimization of automotive components to achieve lightweight structures, functional integration, as well as customized solutions. By reviewing the literature on additive manufacturing processes and material properties, and examining specific component cases, this study investigated strategies for improving part performance via parameter tuning, data refinement, and topology design. The results reveal that additive manufacturing can notably reduce part weight and enhance overall system performance, yet it still faces challenges such as process defects, an incomplete standardization system, and industrial application bottlenecks. Future efforts should prioritize multi-material integrated printing and the intelligent closed-loop management of process-performance, which are key to wider adoption of additive manufacturing in the automotive sector.

Human behavior correct modeling is very essential for next gen intelligent system, but it still has a lot of practical difficulties to be addressed even though it can deal with high dimensional data better than traditional Machine Learning (ML). Multimodal data fusion, loss of long-term temporal information, lack of interpretability. These are the problems. This paper looks into how deep learning is used in three main parts: feeling figuring out, social interaction patterns, and everyday actions watching. These findings reveal a distinct ket-set: structures like Long Short-Term Memories (LSTMs) and Convolutional Neural Netwroks (CNNs), they kasper tasks simple things really good, their ability add the exact randomness catching real societies is harder though. As well as that, Graph Neural Networks (GNNs) can model relationships but have computational scalability issues. Finally, this study recommended using Explainable Artificial Intelligence (XAI) along with privacy preserving method to solve the problem of black box and fill the gap between practical performance and reliable application of XAI.

Against the backdrop of the advancement of educational digitalization, smart sports have become the core direction of physical education teaching reform in colleges and universities. However, the evaluation of gymnastics teaching in colleges and universities is deeply trapped in the traditional model and disconnected from the trend of precise quantification. This paper integrates video stream recognition, skeleton key point evaluation and lightweight CNN technology to design a real-time evaluation algorithm suitable for classroom scenarios. Verified by 12,000 gymnastic movement samples, this algorithm achieves a balance between lightweight and precision, providing technical support for the data-driven transformation of gymnastic teaching.

This research addresses how an interactive digital tool might emotionally connect the youth with their culture, concentrating on one of the oldest and richest traditions of China - ancient music. The application enables the user to be the instrumentalist of the virtual world through playing the guqin, bianzhong and some popular songs as well, all this while getting step-by-step instructions, a scoring feature, a playback feature, and cultural insights presented in the form of trivia tied to high scores. The purpose of the app is to use the emotional and cultural aspects to bring the past into the young people's present time and thus get them more involved. It was the systematic measures of cultural and musical engagement together with emotional reactions to learning and music that I used for the research whose participants were thirty-six young adults aged 18 to 25 and who completed a 14-day study, where the influence of the traditional cultural activities was compared with this interactive gamified method. The data imply that the app users practice the culture more deeply than other users, so interactive digital instruments can become a powerful tool to revive cultural heritage and, at the same time, make it more fun and educational. The research outcome can shape not only the future of cultural preservation but also that of education.

In game theory, there is a fundamental challenge about maintaining cooperation among selfish players, especially under practical noise. This study applies a noisy Iterated Prisoners Dilemma (IPD) model to investigate how learning strategies perform against classical strategies when players may receive false or misleading signals due to random observation errors. More specifically, this study compares Deep Q-Network (DQN) agents with basic Q-learning (QL) and several classical strategies such as Tit-for-Tat, Win-Stay-Lose-Shift, and Grudger. The experiment results show that when noise emerges, DQN agents not only achieve higher cumulative rewards than other strategies but also maintain more stability, adaptability, and resilience across repeated interactions. DQN agents deep neural structure helps them to capture long-term temporal dependencies, effectively differentiate accidental defections from intentional ones, and recover cooperation after disturbances by noise. These findings indicate that deep reinforcement learning is effective in noisy and imperfect settings. The findings also offers valuable insights for understanding the emergence of cooperation and for designing robust multi-agent decision-making mechanisms in noisy or uncertain environments.

With the growing popularity of chatbot programs like ChatGPT and DeepSeek, people are no longer content with merely simple text for the content they generate. With the popularization of the IT and Internet industries, the demand for generating code based on natural language descriptions is increasing day by day. This paper will experimentally demonstrate how fine-tuning the CodeLlama model based on Parameter Efficient fine-tuning (PEFT), especially low-rank adaptive (LoRA) technology, optimizes and improves the generation of Python code, and use authoritative evaluation indicators to prove this point. We will conduct research on code generation by drawing on datasets such as HumanEval and APPs. Then, we will fine-tune the existing baseline model using fine-tuning techniques. We will comprehensively evaluate the generation results before and after fine-tuning using the pass@k metric, plot the results in a chart for comparison, and draw conclusions accordingly. The experimental results show that, on the whole, the Code Llama model before and after LoRA fine-tuning has a significant improvement in the pass rate of the generated code. However, under different indicators, the improvement effect varies slightly. Under the evaluation system of pass@k, the greater the k value, the higher the absolute percentage increase, but the lower the relative percentage increase. Under the condition that other factors remain unchanged, although increasing the rank of the fine-tuning matrix can improve the pass rate of the generated code, the training time will significantly increase, indicating certain limitations.

The intensifying competition among streaming platforms has made precise recommendation a core challenge for enhancing user retention and content value. Traditional collaborative filtering and shallow latent models have limitations in handling data sparsity and modeling nonlinear preferences, making it difficult to effectively characterize user interest structures. This paper proposes an improved neural collaborative filtering method that models the nonlinear interactions between users and movies through embedding representations and multi-layer perceptrons. We construct an implicit feedback training set based on the MovieLens dataset and adopt a negative sampling strategy to enhance preference discrimination. Experimental results show that the model performs outstandingly in Top-N recommendation tasks, with Hit Ratio@10 reaching 0.87, which is significantly better than matrix factorization and the original NCF method. The study indicates that deep interaction modeling can more accurately capture latent preferences, providing an effective path for large-scale personalized recommendations.

Large language models (LLMs) are increasingly used for information extraction, and they often perform well when the target facts are stated plainly. In scientific papers, though, many relationships are only implied: the reader has to connect evidence across sentences, track context, and sometimes make several small inferences before a relation becomes clear. How reliably LLMs handle that kind of reading remains an open question. To study this, we propose an evaluation framework that tests relation extraction under progressively heavier reasoning demands. It relies on three prompt templates that grow more structured and more challenging, and we apply them to a curated perovskite solar cell dataset covering three core categories of scientific informationmaterials, processes, and performance. We then benchmark three representative models. The results show a consistent pattern. When relations are explicit, models extract them with high accuracy. As soon as the task requires deeper inferenceespecially for implicit or multi-hop relations that hinge on nuanced semanticsperformance drops noticeably. The models also fail in different ways: some are comparatively good at maintaining coherent, step-by-step reasoning, while others are more dependable at avoiding hallucinated links or preserving type consistency. Even with these strengths, sizable gaps remain in coverage, reasoning depth, and overall completeness. Taken together, the findings clarify where current LLMs are genuinely useful for scientific information extraction and where they still fall short. They tend to be most reliable in structured extraction settings, while deeper semantic understanding remains the main bottleneck. By quantifying performance across prompt levels, the framework offers a practical yardstick for semantic comprehension and scientific reasoning in relation extraction, and it provides empirical support for downstream applications such as scientific knowledge graph construction.

Rising demand and persistent capacity constraints in mental health care have exposed the limitations of rule-based triage and single-shot trial designs, especially under volatile symptoms, delayed outcome feedback, and fragmented digital data. To address these challenges, this study develops a reinforcement learningdriven counterfactual policy evaluation framework that models care allocation as a partially observable Markov decision process and combines doubly robust off-policy estimation with conservative policy improvement. In-hospital follow-up trajectories and community wearable data are integrated to construct latent state representations, and candidate policies for visit frequency, session duration, and intervention modality are evaluated offline for value and safety before deployment. On real-world cohorts, the doubly robust estimator achieves an 8.3% improvement in policy value over the baseline while retaining 76.2% of the original effective sample size and reducing bias and mean squared error under optimized clipping. Operationally, the learned policy shortens median waiting time to 8.7 days, increases per-staff throughput by 22.6%, and lowers readmission rates, while opportunity and treatment intensity gaps shrink and stability metrics improve. These findings indicate that RL combined with counterfactual evaluation can support more efficient, fair, and auditable dynamic allocation of mental health services under fixed resources and offer a transferable pipeline from offline data to safe policy deployment.

Massive Multiple-Input Multiple-Output (Massive MIMO) serves as a foundational enabling technology for 5G and future communication systems, markedly boosting spectral and energy efficiency through the deployment of large-scale antenna arrays. However, the scaling-up of antenna arrays has led to a substantial increase in system power consumption and hardware costs, with high-precision analog-to-digital converters (ADCs) emerging as the dominant power consumption bottleneck in the radio frequency chain. To alleviate system complexity and power consumption, low-resolution ADCs (13 bits) have attracted extensive research interest in recent years. Such schemes can substantially curtail hardware costs and energy consumption while retaining satisfactory system performance. Nevertheless, the introduction of severe nonlinear distortion due to low-precision quantization disrupts the linear Gaussian model assumption upon which traditional receiver algorithms rely, resulting in compromised channel estimation and signal detection performance. Quantization errors demonstrate non-Gaussian and input-dependent characteristics, leading to the degradation of amplitude information and thus constraining the applicability of technologies such as high-order modulation and high-precision sensing. This paper presents a systematic review of low-precision quantization techniques for Massive MIMO. It first investigates the impacts of low-bit quantization on system models and signal statistical properties. Subsequently, it elaborates on transceiver architectures and key design challenges pertaining to low-precision ADCs/DACs. The paper highlights signal processing and algorithmic strategies to overcome quantization distortion, including Bussgang decomposition linearization methods, statistical inference techniques such as approximate message passing (AMP), model-driven deep learning frameworks, and quantization architectures endowed with noise-shaping capabilities. Finally, it discusses the challenges and future directions of this technology in emerging scenarios, including terahertz communications, intelligent reflecting surfaces, and integrated sensing and communication. This paper seeks to provide researchers with a systematic technical overview, clarifying the intrinsic connections and trade-offs among different methods, and offering valuable insights for the realization of high-energy-efficiency and low-cost Massive MIMO systems.