Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
838
result(s) for
"Huang, Yanjun"
Sort by:
Shapley value: from cooperative game to explainable artificial intelligence
With the tremendous success of machine learning (ML), concerns about their black-box nature have grown. The issue of interpretability affects trust in ML systems and raises ethical concerns such as algorithmic bias. In recent years, the feature attribution explanation method based on Shapley value has become the mainstream explainable artificial intelligence approach for explaining ML models. This paper provides a comprehensive overview of Shapley value-based attribution methods. We begin by outlining the foundational theory of Shapley value rooted in cooperative game theory and discussing its desirable properties. To enhance comprehension and aid in identifying relevant algorithms, we propose a comprehensive classification framework for existing Shapley value-based feature attribution methods from three dimensions: Shapley value type, feature replacement method, and approximation method. Furthermore, we emphasize the practical application of the Shapley value at different stages of ML model development, encompassing pre-modeling, modeling, and post-modeling phases. Finally, this work summarizes the limitations associated with the Shapley value and discusses potential directions for future research.
Journal Article
Safety of Autonomous Vehicles
Autonomous vehicle (AV) is regarded as the ultimate solution to future automotive engineering; however, safety still remains the key challenge for the development and commercialization of the AVs. Therefore, a comprehensive understanding of the development status of AVs and reported accidents is becoming urgent. In this article, the levels of automation are reviewed according to the role of the automated system in the autonomous driving process, which will affect the frequency of the disengagements and accidents when driving in autonomous modes. Additionally, the public on-road AV accident reports are statistically analyzed. The results show that over 3.7 million miles have been tested for AVs by various manufacturers from 2014 to 2018. The AVs are frequently taken over by drivers if they deem necessary, and the disengagement frequency varies significantly from 2 × 10−4 to 3 disengagements per mile for different manufacturers. In addition, 128 accidents in 2014–2018 are studied, and about 63% of the total accidents are caused in autonomous mode. A small fraction of the total accidents (∼6%) is directly related to the AVs, while 94% of the accidents are passively initiated by the other parties, including pedestrians, cyclists, motorcycles, and conventional vehicles. These safety risks identified during on-road testing, represented by disengagements and actual accidents, indicate that the passive accidents which are caused by other road users are the majority. The capability of AVs to alert and avoid safety risks caused by the other parties and to make safe decisions to prevent possible fatal accidents would significantly improve the safety of AVs. Practical applications. This literature review summarizes the safety-related issues for AVs by theoretical analysis of the AV systems and statistical investigation of the disengagement and accident reports for on-road testing, and the findings will help inform future research efforts for AV developments.
Journal Article
SURVIVAL AND DEVELOPMENT: RESEARCH ON CHINESE HIGHER EDUCATION DURING THE WAR OF RESISTANCE AGAINST JAPANESE AGGRESSION (1931-1945)
China’s higher education system demonstrated remarkable resilience during the War of Resistance Against Japanese Aggression (1931–1945), a period marked by national crisis, instability, and widespread destruction. Universities and colleges were forced to evacuate from war-torn regions, often relocating to remote and underdeveloped areas in the country’s interior. Despite facing immense logistical and resource-related obstacles, these institutions continued to function, embodying the spirit of Higher Education Resilience. This study investigates how Chinese universities adapted under wartime conditions, focusing specifically on how they preserved Educational Quality, aligned with National Defense, and maintained Intellectual Independence in the face of adversity. Drawing on 17 semi-structured interviews with scholars, educators, and individuals closely connected to this historical period, the research employs a qualitative methodology grounded in thematic analysis. Th e three-step coding process revealed critical themes that capture the multifaceted response of higher education institutions during the war. Among the most significant were the adaptability of academic administration, innovation in curriculum design to support National Defense Alignment, and the maintenance of academic rigor despite the scarcity of materials and infrastructure. A notable outcome of university relocation was the unintentional stimulation of regional development, as academic institutions brought intellectual resources and cultural vitality to inland provinces. Participants also emphasized that universities served as crucial sites for sustaining Chinese cultural identity, promoting national unity, and resisting intell ectual repression, thereby safeguarding Intellectual Independence. Despite the physical and psychological toll of war, faculty and students persevered, contributing to both the war effort and the preservation of knowledge. This research contributes new perspectives to the discourse on crisis-era education by highlighting how higher education can remain robust even in extreme conditions. The findings illustrate that Higher Education Resilience is not merely institutional survival, but also the capac ity to uphold core educational values and respond to national needs. Ultimately, the study underscores that universities play a vital role in national recovery and development, particularly in times of profound turmoil.
Journal Article
Antibacterial Efficacy and Mechanisms of Curcumin-Based Photodynamic Treatment against Staphylococcus aureus and Its Application in Juices
Antimicrobial Photodynamic Treatment (aPDT) is a non-thermal sterilization technology, which can inactivate common foodborne pathogens. In the present study, photodynamic inactivation on Staphylococcus aureus (S. aureus) with different concentrations of curcumin and light dose was evaluated and the mechanisms were also investigated. The results showed that curcumin-based aPDT could inactivate S. aureus cells by 6.9 log CFU/mL in phosphate buffered saline (PBS). Moreover, the modified Gompertz model presented a good fit at the inactivation data of S. aureus. Photodynamic treatment caused cell membrane damage as revealed by analyzing scanning electron microscopy (SEM) images. Leakage of intracellular constituents further indicated that cell membrane permeability was changed. Flow cytometry with double staining demonstrated that cell membrane integrity and the activity of nonspecific esterase were destroyed. Compared with the control group, intracellular reactive oxygen species (ROS) levels caused by photodynamic treatment significantly increased. Furthermore, curcumin-based aPDT reduced S. aureus by 5 log CFU/mL in juices. The color of the juices was also tested using a Chromatic meter, and it was found that b* values were the most markedly influenced by photodynamic treatment. Overall, curcumin-based aPDT had strong antibacterial activity against S. aureus. This approach has the potential to remove foodborne pathogens from liquid food.
Journal Article
Early prediction of lithium-ion battery degradation with a generative pre-trained transformer
The early detection of degradation in lithium-ion batteries (LIBs) is crucial for effective predictive maintenance and recycling. However, accurately predicting the future degradation of LIBs in early stage is challenging due to the barely noticeable performance changes at initial charging cycles and the long-term nonlinear degradation pattern. In this work, we propose a two-stage early-stage degradation prediction method, BatteryGPT, which employs a Generative Pre-trained Transformer (GPT) to autoregressively predict the charging data of entire lifecycle and a state-of-health (SOH) estimator to correlates the predicted charging data with ageing features in LIBs. The validation demonstrates that BatteryGPT can predict the future LIB degradation with high accuracy using early charging data, before any capacity degradation is evident. Predicting with the first 30% of the battery lifetime, BatteryGPT significantly outperforms baselines, achieving a root mean square error (RMSE) of 0.213% for SOH variation prediction, and mean absolute percent errors (MAPE) of 2.30% and 1.18% for knee point and EOL predictions. Even predicting with the first 5% of lifetime charging data, BatteryGPT demonstrates strong early-stage prediction performance.
The early detection of degradation in lithium-ion batteries is crucial for effective predictive maintenance and recycling. Here, the authors propose a two-stage early-stage degradation prediction method based on GPT to predict the charging data of and state-of-health of batteries
Journal Article
Adsorption Performance of Methylene Blue by KOH/FeCl3 Modified Biochar/Alginate Composite Beads Derived from Agricultural Waste
In this study, high-performance modified biochar/alginate composite bead (MCB/ALG) adsorbents were prepared from recycled agricultural waste corncobs by a high-temperature pyrolysis and KOH/FeCl3 activation process. The prepared MCB/ALG beads were tested for the adsorption of methylene blue (MB) dye from wastewater. A variety of analytical methods, such as SEM, BET, FTIR and XRD, were used to investigate the structure and properties of the as-prepared adsorbents. The effects of solution pH, time, initial MB concentration and adsorption temperature on the adsorption performance of MCB/ALG beads were discussed in detail. The results showed that the adsorption equilibrium of MB dye was consistent with the Langmuir isothermal model and the pseudo-second-order kinetic model. The maximum adsorption capacity of MCB/ALG−1 could reach 1373.49 mg/g at 303 K. The thermodynamic studies implied endothermic and spontaneous properties of the adsorption system. This high adsorption performance of MCB/ALG was mainly attributed to pore filling, hydrogen bonding and electrostatic interactions. The regeneration experiments showed that the removal rate of MB could still reach 85% even after five cycles of experiments, indicating that MCB/ALG had good reusability and stability. These results suggested that a win-win strategy of applying agricultural waste to water remediation was feasible.
Journal Article
Prognostic impact of lymphovascular and perineural invasion in squamous cell carcinoma of the tongue
This study aimed to investigate the prognostic impact of lymphovascular and perineural invasions in patients with squamous cell carcinoma of the tongue who received surgery-based treatment at our institution between January 2013 and December 2020. Patients were divided into four groups based on the presence of perineural (P−/P +) and lymphovascular invasions (V−/V +): P–V−, P–V + , P + V−, and P + V + . Log-rank and Cox proportional hazard models were used to evaluate the association between perineural /lymphovascular invasion and overall survival (OS). Altogether, 127 patients were included, and 95 (74.8%), 8 (6.3%), 18 (14.2%), and 6 (4.7%) cases were classified as P–V−, P–V + , P + V−, and P + V + , respectively. Pathologic N stage (pN stage), tumor stage, histological grade, lymphovascular invasion, perineural invasion, and postoperative radiotherapy were significantly associated with OS (p < 0.05). OS was significantly different among the four groups (p < 0.05). Significant between-group differences in OS were detected for node-positive (p < 0.05) and stage III–IV (p < 0.05) cases. OS was the worst in the P + V + group. Lymphovascular and perineural invasions are independent negative prognostic factors for squamous cell carcinoma of the tongue. Patients with lymphovascular and/or perineural invasion may have significantly poorer overall survival than those without neurovascular involvement.
Journal Article
Aptamer-based fluorescent screening assay for acetamiprid via inner filter effect of gold nanoparticles on the fluorescence of CdTe quantum dots
This paper reports a novel aptamer-based fluorescent detection method for small molecules represented by acetamiprid based on the specific binding of aptamers with acetamiprid, and the inner filter effect (IFE) of gold nanoparticles (AuNPs) on the fluorescence of CdTe quantum dots (CdTe QDs). When CdTe QDs were mixed with AuNPs, the fluorescence of CdTe QDs was significantly quenched via IFE. The IFE efficiency could be readily modulated by the absorption and the aggregation state of AuNPs. The presence of salt could easily induce the aggregation of AuNPs, resulting in the fluorescence recovery of the quenched QDs. Acetamiprid-binding aptamer (ABA) could adsorb on the negatively charged AuNPs through the coordination interaction to protect AuNPs from salt-induced aggregation, so the fluorescence of CdTe QDs would be quenched by the IFE of AuNPs. However, the specific binding of ABA with acetamiprid could release the ABA from the surfaces of AuNPs and decrease the salt tolerance of AuNPs, so the IFE-decreased fluorescence of CdTe QDs was regained with the presence of acetamiprid, and the fluorescence enhancement efficiency was driven by the concentration of acetamiprid. Based on this principle, the aptamer-based fluorescent method for acetamiprid has been established and optimized. The assay exhibited excellent selectivity towards acetamiprid over its analogues and other pesticides which may coexist with acetamiprid. Under the optimum experiment conditions, the established method could be applied for the determination of acetamiprid with a wide linear range from 0.05 to 1.0 μM, and a low detection limit of 7.29 nM (3σ). Furthermore, this IFE-based method has been successfully utilized to detect acetamiprid in six types of vegetables, and the results were in full agreement with those from HPLC and LC-MS. The proposed method displays remarkable advantages of high sensitivity, rapid analysis, excellent selectivity, and would be suitable for the practical application of target screening in real samples. Graphical Abstract ᅟ
Journal Article
Effective Removal of Fe (III) from Strongly Acidic Wastewater by Pyridine-Modified Chitosan: Synthesis, Efficiency, and Mechanism
A novel pyridine-modified chitosan (PYCS) adsorbent was prepared in a multistep procedure including the successive grafting of 2-(chloromethyl) pyridine hydrochloride and crosslinking with glutaraldehyde. Then, the as-prepared materials were used as adsorbents for the removal of metal ions from acidic wastewater. Batch adsorption experiments were carried out to study the impact of various factors such as solution pH value, contact time, temperature, and Fe (III) concentration. The results showed that the absorbent exhibited a high capacity of Fe (III) and the maximum adsorption capacity was up to 66.20 mg/g under optimal experimental conditions (the adsorption time = 12 h, pH = 2.5, and T = 303 K). Adsorption kinetics and isotherm data were accurately described by the pseudo-second-order kinetic model and Sips model, respectively. Thermodynamic studies confirmed that the adsorption was a spontaneous endothermic process. Moreover, the adsorption mechanism was investigated using Fourier transform infrared spectroscopy (FTIR) and X-ray photoelectron spectroscopy (XPS). The results revealed the pyridine group forms a stable chelate with iron (III) ions. Therefore, this acid-resistant adsorbent exhibited excellent adsorption performance for heavy metal ions from acidic wastewater compared to the conventional adsorbents, helping realize direct decontamination and secondary utilization.
Journal Article