Effects of hydration parameters on chemical properties of biocrudes based on machine learning and experiments

To investigate the effects of temperature and biomass concentration of Hydrothermal liquefaction (HTL) on chemical properties of biocrudes, machine learning (ML) was used to predict the weight of hydration parameters on the properties of biocrudes. The elemental compositions, molecular weights, functional groups, thermal degradation, molecular structure of biocrudes were studied. The optimum yield of biocrudes was 65% and the highest heat value reached up to 34.28 kJ/g, showing comparable fuel properties. It was found that the hydration temperature significantly affects the elemental components, functional groups and molecular weight and structures of biocrudes. In addition, biomass concentration also affect the functional groups and structures of biocrudes. ML results indicated that Support Vector Machine Linear Kernel method is suitable for heat value prediction.

Application of deep learning in image recognition and diagnosis of gastric cancer

In recent years, artificial intelligence has been extensively applied in the diagnosis of gastric cancer based on medical imaging. In particular, using deep learning as one of the mainstream approaches in image processing has made remarkable progress. In this paper, we also provide a comprehensive literature survey using four electronic databases, PubMed, EMBASE, Web of Science, and Cochrane. The literature search is performed until November 2020. This article provides a summary of the existing algorithm of image recognition, reviews the available datasets used in gastric cancer diagnosis and the current trends in applications of deep learning theory in image recognition of gastric cancer. covers the theory of deep learning on endoscopic image recognition. We further evaluate the advantages and disadvantages of the current algorithms and summarize the characteristics of the existing image datasets, then combined with the latest progress in deep learning theory, and propose suggestions on the applications of optimization algorithms. Based on the existing research and application, the label, quantity, size, resolutions, and other aspects of the image dataset are also discussed. The future developments of this field are analyzed from two perspectives including algorithm optimization and data support, aiming to improve the diagnosis accuracy and reduce the risk of misdiagnosis.

A matrix effect correction method for fissile nuclear material mass measurement by delayed neutrons

Active neutron interrogation (ANI), extensively employed in nuclear safeguards, is sensitive to the presence of special nuclear materials (SNMs). However, the interaction of the matrix material with neutrons weakens the precision of fissile mass measurements by the ANI system. Therefore, it is paramount to ensure that the evaluation of the fissile mass remains unaffected by the matrix and enhances the assay performance of the ANI system. The present work proposes a matrix correction method based on the neutron flux monitor group (NFMG) response to tackle this issue. Based on the varying influence of different matrices on the neutron energies and fluxes, the NFMG response can be used to quantify the matrix effects. This allows the method to enable the identification of matrices already present in the database and has the potential to compensate for unknown matrix effects. To validate the applicability and accuracy of this method, a Shuffler system model and various matrix compositions were developed using the Geant4 toolkit. The results demonstrate that the improved simulated annealing (SA) algorithm exhibits excellent stability when confronted with varying enrichments and distributions of U3O8 materials. When the threshold is set at α ≥ 0.90, the improved SA algorithm achieves a successful identification rate of 90.4% for matrices. Simultaneously, using the response equation, the average relative deviation of the corrected 235U mass is no more than 7%. For the unknown matrix, the correction capability of this method relies primarily on the construction of the reference database and the response equation. In most cases, the average relative deviation of the corrected 235U mass is less than 28%.

Optimization of Forward Collision Warning Algorithm Considering Truck Driver Response Behavior Characteristics

Forward collision warning (FCW) systems have been widely used in trucks to alert drivers of potential road situations so they can reduce the risk of crashes. Research on FCW use shows, however, that there are differences in drivers' responses to FCW alerts under different scenarios. Existing FCW algorithms do not take differences in driver response behavior into account, with the consequence that the algorithms' minimum safe distance assessments that trigger the warnings are not always appropriate for every driver or situation. To reduce false alarms, this study analyzed truck driver behavior in response to FCW warnings, and k-means clustering was adopted to classify driver response behavior into three categories: Response Before Warning (RBW), Response After Warning (RAW), and No Response (NR). Results showed that RBW clusters tend to occur at long following distances (>19 m), and drivers applied braking before the warning. In RAW clusters, deceleration after warning is significantly more forceful than before warning. NR clusters occur at short distances, and deceleration fluctuates only slightly. To optimize the FCW algorithm, the warning distance was divided into reaction distance and braking distance. The linear support vector machine was used to fit the driver reaction distance. The long short-term memory method was used to predict braking distance based on each of the three response scenarios: R2 was 0.896 for RAW scenarios, 0.927 for RBW scenarios, and 0.980 for NR scenarios. Verification results show that the optimized truck FCW algorithm improved safety by 1 % to 5.1 %; accuracy reached 97.92 %, and the false alarm rate was 1.73 %.

Establishment of an NPK nutrient monitor system in yield-graded cotton petioles under drip irrigation

BACKGROUND

The determination of nutrient content in the petiole is one of the important methods for achieving cotton fertilization management. The establishment of a monitoring system for the nutrient content of cotton petioles during important growth periods under drip irrigation is of great significance for achieving precise fertilization and environmental protection.

METHODS

A total of 100 cotton fields with an annual yield of 4500-7500 kg/ha were selected among the main cotton-growing areas of Northern Xinjiang. The nitrate nitrogen (NO3--N), inorganic phosphorus (PO43--P) and inorganic potassium (K+-K) content and yield of cotton petioles were recorded. Based on a yield of 6000 kg/ha as the dividing line, a two-level and yield-graded monitoring system for NO3--N, PO43--P and K+-K in cotton petioles during important growth periods was established, and predictive yield models for NO3--N, PO43--P and K+-K in petioles during important growth periods were established.

RESULTS

The results showed found that the yields of the 100 cotton fields surveyed were normally distributed. Therefore, two yield grades were classified using 6000 kg/ha as a criterion. Under different yield-graded, the NO3--N, PO43--P and K+-K content of petiole at important growth stages was significantly positively correlated with yield. Further, the variation range of NO3--N, PO43--P and K+-K content in petioles could be used as a standard for yield-graded. In addition, a yield prediction model for the NO3--N, PO43--P and K+-K content of petioles was developed. The SSO-BP validation model performed the best (R2 = 0.96, RMSE = 0.06 t/ha, MAE = 0.05 t/ha) in the full bud stage, which was 12.9% higher than the BP validation model. However, the RMSE and MAE were decreased by 86.7% and 88.1%, respectively.

CONCLUSION

The establishment of NPK nutrition monitor system of cotton petioles under drip irrigation based on yield-graded provides an important basis for nutrition monitor of cotton petiole under drip irrigation in Xinjiang. It also provides a new method for cotton yield prediction.

Enhanced prediction of total purine content in hyperspectral images of diverse livestock meat samples using optimization algorithm

This research looks at rapid detection and algorithm optimization of total purine in livestock meat with the goal of supporting the creation of low-purine diets. First, the prediction models for total purine content of single species were constructed based on chemical and hyperspectral data. Next, the correlation of spectral curves between different species by Hausdorff distance and Pearson correlation coefficient were analyzed to explore differential purine variation in mixed livestock meat. Finally, the optimal prediction models of total purines in mixed livestock meat were obtained, which were the interval-variable-iterative-spatial-contraction-method-sparrow-search-algorithm-bidirectional-long-short-term-memory-mmulti-head-attention (SSA-Bi-LSTM-MHA) in visible near-infrared hyperspectral imaging technology (Vis-NIR HSI) and iteratively-retain-informative-variables-SSA-Bi-LSTM-MHA in NIR-HSI, respectively, with Rp2 of 0.7820 and 0.7766. Totally, the conclusion that model performance of mixed samples due to increased computational complexity was lower than the ideal model of single samples was obtained and validated. This study demonstrated the potential of HSI for rapid detection of purine content, which further promoted the industrialization of online monitoring of livestock meat quality.

Performance evaluation of air-source heat pump based on a pressure drop embedded model

An air-source heat pump simulation model, accounting for evaporator and condenser pressure drop, has been developed. The model is capable of computing the heat pump's coefficient of performance (COP) under different ambient temperatures and relative humidities above frosting conditions. This research extends an existing iterative simulation method that relies on the equalization of logarithmic mean temperature differences (LMTDs) calculated through two different approaches by adding a pressure drop simulation. Frictional and acceleration pressure drop is considered, computed iteratively. Simulation results for three different refrigerants, R410A, R32 and R290, are compared. The model's accuracy is validated by comparing simulated COP values with measured COP values from the reference heat pump datasheet. The model closely replicates the measured COP values above frosting conditions, with only a slight underestimation of approximately 1.5%. Results show a substantial impact of ambient temperature on the COP. For instance, an ambient temperature of 20 ◦C, compared to 7 ◦C, results in a COP increase of up to 35%, while an ambient temperature of -10 ◦C leads to a 26% reduction in COP. Relative humidity enhances the COP if air moisture condensation becomes possible. Higher condenser capacities negatively affect the COP. The study highlights the differences in pressure drop characteristics between the condenser and the evaporator for the modeled heat pump, with maximum pressure drops of 220 kPa and 50 kPa for the condenser and evaporator, respectively. Additionally, the choice of refrigerant significantly influences pressure drop, with R32 displaying the lowest pressure drop, R410A showing the highest condenser pressure drop, and R290 causing the highest evaporator pressure drop.

The analysis of optimization in music aesthetic education under artificial intelligence

In the artificial intelligence (AI) domain, effectively integrating deep learning (DL) technology with the content, teaching methodologies, and learning processes of music aesthetic education remains a subject worthy of in-depth exploration and discussion. The aim is to meet to the music aesthetic needs of students across different age groups and levels of musical literacy. In this paper, the concepts of AI and DL algorithm are first introduced, and their algorithm principles and application status are understood. Then, they are integrated into the application of music aesthetic education, and the algorithm principles and running codes are designed. Finally, experiments are carried out to verify the accuracy of music emotion recognition based on DL algorithm in AI environment to verify the effectiveness of music aesthetic education method based on DL. The results show that the algorithm proposed in this paper has higher accuracy, which combines the advantages of AI and DL algorithm, and obtains higher recognition accuracy. It provides more possibilities for future music aesthetic teaching activities. This paper is dedicated to investigating the feasibility and approach to optimizing the method of music aesthetic education through DL. Its objective is to chart a new developmental direction and practical pathway for music aesthetic education in the era of AI.

Optimizing the compliance third-party supervision workflow of involved enterprises using artificial intelligence ant colony algorithm

This article optimized the compliance third-party supervision workflow of the involved enterprises based on the artificial intelligence ant colony optimization (ACO) algorithm. The basic principles and application advantages of ACO were introduced, and a heuristic information matrix was defined using ACO to optimize the data collection and analysis stage of the compliance third-party supervision workflow. During the experimental phase, a feasibility analysis was conducted on the optimization of third-party supervision workflows for compliance by ACO involved enterprises through simulation experiments. The experiments were evaluated from four aspects: data quality, model performance, scheme effectiveness, and supervision effectiveness. Among the metrics data for the ACO-optimized test set were 0.03 and 0.025 for MSE (Mean Square Error) and Γ, 0.8, 0.78, 0.79, and 0.88 for Accuracy, Recall, F1 Score, and AUC-ROC (Area Under the Curve-Receiver Operating Characteristic), and 0.28, 0.4, 0.88, and 0.12 for CER (Cost-Effectiveness Ratio), NPV (Net Present Value), SCR (Supervision Coverage Rate), and CRC (Compliance Rate Change), respectively. The experimental results showed that, in terms of data quality, model performance, scheme effectiveness, and supervision effectiveness, the evaluation indicators of the compliance third-party supervision workflow of the involved enterprises optimized using ACO were superior to those without ACO optimization.

Differences in emotional expression among college students: a study on integrating psychometric methods and algorithm optimization

BACKGROUND

College students are in an important stage of life development, and their emotional expression ability has a profound impact on their mental health, interpersonal relationships, and academic performance. There are significant differences in emotional expression among individuals, which are influenced by various factors such as gender, cultural background, and personality traits. However, traditional research on emotional expression often relies on a single measurement method, which has problems such as single data dimensions, limited analysis methods, and lack of real-time dynamism and personalization. To overcome these limitations, this study conducted a comprehensive analysis using psychometric methods and algorithm optimization techniques.

METHODS

The Emotional Intelligence Scale (EQ-i) and the depression-anxiety-stress-21 (DASS-21) were used to quantitatively evaluate the emotional state of college students, and their facial expressions and speech emotion data were collected. In order to improve the precision of data analysis, random forests, support vector machines, and neural network machine learning algorithms were applied, and the variance analysis was used to calculate and compare the emotional differences of different genders and academic backgrounds in different grades.

RESULTS

The research results showed that gender, major, and grade differences significantly affected the emotional expression of college students. The F-values for the total EQ-i score of different genders were 7.00, and the F-values for depression, anxiety, and stress scores between different grades were 22.45, 12.48, and 9.14. Male engineering students scored higher in emotional intelligence than female liberal arts students, but liberal arts students showed more significant improvement in later academic years, reflecting the differing impacts of disciplinary environments on emotional development. Female students generally exhibited higher levels of anxiety and stress, particularly those in liberal arts, while female engineering students faced additional psychological burdens due to gender imbalance and biases. Anxiety and stress levels increased across all students as they advanced in their studies, correlating with academic and graduation pressures.

CONCLUSION

This article was based on the integration of psychometric methods and algorithm optimization techniques, exploring the differences in emotional expression among college students, providing new ideas for personalized mental health interventions for college students, enriching the theoretical basis of emotional expression research, and providing important references for education and mental health practice.

A novel hybrid machine learning approach for accurate retrieval of ocean surface chlorophyll-a across oligotrophic to eutrophic waters

Accurate assessment of chlorophyll a (Chla) concentration distribution and variations is significant for environmental monitoring and ecological research. However, the inversion of Chla in different optical types of water bodies can only be achieved by establishing algorithms suitable for different optical types, lacking a machine learning algorithm framework. Therefore, this study focuses on two aspects, input features and data samples, and designs an innovative composite machine learning algorithm framework called Synth Ridge Framework (SRF). The framework mainly consists of two main components: feature expansion and model construction. We employed the band ratio method and BorutaShap for feature expansion and selection. By integrating three gradient boosting decision tree models (XGBoost, LightBoost, and CatBoost) with the MDN ensemble strategy, we constructed a model named SynthRidge, aiming to enhance the model's overall performance. SynthRidge was trained and validated using the Rrs-In situ Chla dataset from the Terra-MODIS sensor, with Chla values ranging from 0 to 50 mg/m3 in both datasets. On mg/m3the validation dataset, the SynthRidge model achieved strong predictive performance, with an R2 of 0.930, a slope of 0.928, an RMSE of 4.672 mg/m3, an RMLSE of 0.039, a bias of 1.023, and an MAE of 1.389. Compared to the best-performing baseline model, the GBDT ensemble, SynthRidge demonstrated superior accuracy and robustness. Specifically, it improved the R2 by 0.006, increased the slope by 0.020, reduced the RMSE by 0.890 mg/m3, and decreased the RMLSE by 0.003. The validation dataset has its R2, Slope, RMSE, RMLSE, Bias, and MAE values of 0.930, 0.928, 4.672 mg/m3, 0.039, 1.023, and 1.389, respectively. The predicted Chla density distribution by SynthRidge was more consistent with the measured values. These findings suggest that SRF is capable of effectively compensating for the limitations of input features, reducing the negative impact of data distribution, and improving the limitations of complex fusion algorithms. Furthermore, the performance of SRF on the SeaWiFS dataset demonstrates its versatility across different sensors.