Random Neural Network Model Research Articles

Climb Performance Prediction in High Drag Configuration Middle-Class Transportation Aircraft: An Ensemble Learning Approach

This study addresses the application of machine learning and artificial neural network models for predicting the climb speed of the C-130H military transport aircraft. Random Forest, Neural Network, and Ensemble models were developed to overcome limitations of traditional chart reading and interpolation methods. Models were trained on flight manual data, considering factors such as gross weight, pressure altitude, drag index, temperature deviation, and engine efficiency. Comparative analysis revealed the Ensemble approach, combining Random Forest and Neural Network techniques, provided the highest accuracy (R² ≈ 0.4532), followed by Random Forest (R² ≈ 0.4303) and Neural Network (R² ≈ 0.3765) models. All significantly outperformed the traditional Young Method (R² = -1.2673). Feature importance analysis identified pressure altitude, gross weight, and engine efficiency as critical factors influencing climb speed. The ensemble approach demonstrated more reliable and accurate results in predicting C-130H climb rates, reducing risks associated with single-model reliance. This research highlights the potential of machine learning in aircraft performance prediction, offering possibilities for improving pre-flight preparation, reducing workload, and enhancing flight safety. Implications for the aviation industry and future research directions are discussed, emphasizing the role of advanced predictive models in shaping future flight operations and aircraft performance management.

The Geometry of Flow: Advancing Predictions of River Geometry With Multi‐Model Machine Learning

AbstractHydraulic geometry parameters describing river hydrogeomorphic relationships are critical for determining a channel's capacity to convey water and sediment which is important for flood forecasting. Although well‐established, power‐law hydraulic geometry curves have been widely used to understand riverine systems and mapping flooding inundation worldwide for the past 70 years, we have become increasingly aware of their limitations. In the present study, we have moved beyond these traditional power‐law relationships, testing the ability of machine‐learning models to provide improved predictions of river width and depth. For this work, we have used an unprecedentedly large river measurement data set (HYDRoSWOT) as well as a suite of watershed predictor data to develop novel data‐driven approaches to better estimate river geometries over the contiguous United States (CONUS). Our Random Forest, XGBoost, and neural network models out‐performed the traditional, regionalized power law‐based hydraulic geometry equations for both width and depth, providing R‐squared values of as high as 0.75 for width and as high as 0.67 for depth, compared with R‐squared values of 0.45 for width and 0.18 for depth from the regional hydraulic geometry equations. Our results also show diverse performance outcomes across stream orders and geographical regions for the different machine‐learning models, demonstrating the value of using multi‐model approaches to maximize the predictability of river geometry. The developed models have been used to create the newly publicly available STREAM‐geo data set, which provides river width, depth, width/depth ratio, and river and stream surface area (%RSSA) for nearly 2.7 million NHDPlus stream reaches across the contiguous US.

Predicting stroke risk: An effective stroke prediction model based on neural networks

BackgroundStroke is the leading worldwide cause of disability and death. Effective stroke prevention and management depend on early identification of stroke risk. MethodsEight machine learning algorithms are applied to predict stroke risk using a well-curated dataset with pertinent clinical information. This paper describes a thorough investigation of stroke prediction using various machine learning methods. ResultsThe empirical evaluation yields encouraging results, with the logistic regression, support vector machine, and K-nearest neighbors models achieving an impressive accuracy of 95.04%, and the random forest and neural network models scoring even better, with accuracies of 95.10% and 95.16%, respectively. The neural network exhibits slightly superior performance, indicating its potential as a reliable model for stroke risk assessment. ConclusionsThe empirical evaluation underscores the ability of neural networks to discern intricate data relationships. These findings offer valuable insights for healthcare professionals and researchers, aiding in the development of improved stroke prevention strategies and timely interventions, ultimately enhancing patient outcomes.

Prediction of Hard Drive Failure using Machine Learning

The reliability of hard drives is paramount for maintaining data integrity and availability in cloud services and enterprise-level data centers where unexpected failures significantly impact operational efficiency and general performance. This work aims to develop a predictive model using regression analysis to accurately forecast imminent hard drive failures based on historical operational data, specifically SMART (Self-Monitoring Analysis and Reporting Technology) attributes. The study evaluated various regression models which comprises Decision Tree, Random Forest, Support Vector Machine (SVM), Gradient Boosting, and Neural Network. The outcomes indicated that the Random Forest model, with an MSE of 24.7427 and an R2 of 0.9876 and the Neural Network model, with an MSE of 22.6011 and an R2of 0.7442, as the best performing models as they demonstrated high predictive accuracy and robustness. In contrast, the SVM model showed poor performance with an MSE of 2888.8623 and a negative R2of -0.4465. Based on these outcomes, the Random Forest and Neural Network models are recommended for predicting hard drive failures as they delivered a balance of accuracy and interpretability.

One-way ticket to the moon? An NLP-based insight on the phenomenon of small-scale neo-broker trading

We present an Natural Language Processing based analysis on the phenomenon of “Meme Stocks”, which has emerged as a result of the proliferation of neo-brokers like Robinhood and the massive increase in the number of small-scale stock investors. Such investors often use specific Social Media channels to share short-term investment decisions and strategies, resulting in partial collusion and planning of investment decisions. The impact of online communities on the stock prices of affected companies has been considerable in the short term. This paper has two objectives. Firstly, we chronologically model the discourse on the most prominent platforms. Secondly, we examine the potential for using collaboratively made investment decisions as a means to assist in the selection of potential investments.. To understand the investment decision-making processes of small-scale investors, we analyze data from Social Media platforms like Reddit, Stocktwits and Seeking Alpha. Our methodology combines Sentiment Analysis and Topic Modelling. Sentiment Analysis is conducted using VADER and a fine-tuned BERT model. For Topic Modelling, we utilize LDA, NMF and the state-of-the-art BERTopic. We identify the topics and shapes of discussions over time and evaluate the potential for leveraging information of the decision-making process of investors for trading choices. We utilize Random Forest and Neural Network Models to show that latent information in discussions can be exploited for trend prediction of stocks affected by Social Network driven herd behavior. Our findings provide valuable insights into content and sentiment of discussions and are a vehicle to improve efficient trading decisions for stocks affected from short-term herd behavior.

Leveraging artificial intelligence to identify high-risk patients for postoperative sore throat: An observational study.

Postoperative sore throat (POST) is a prevalent complication after general anesthesia and targeting high-risk patients helps in its prevention. This study developed and validated a machine learning model to predict POST. A total number of 834 patients who underwent general anesthesia with endotracheal intubation were included in this study. Data from a cohort of 685 patients was used for model development and validation, while a cohort of 149 patients served for external validation. The prediction performance of random forest (RF), neural network (NN), and extreme gradient boosting (XGBoost) models was compared using comprehensive performance metrics. The Local Interpretable Model-Agnostic Explanations (LIME) methods elucidated the best-performing model. POST incidences across training, validation, and testing cohorts were 41.7%, 38.4%, and 36.2%, respectively. Five predictors were age, sex, endotracheal tube cuff pressure, endotracheal tube insertion depth, and the time interval between extubation and the first drinking of water after extubation. After incorporating these variables, the NN model demonstrated superior generalization capabilities in predicting POST when compared to the XGBoost and RF models in external validation, achieving an area under the receiver operating characteristic curve (AUROC) of 0.81 (95% CI 0.74-0.89) and a precision-recall curve (AUPRC) of 0.77 (95% CI 0.66-0.86). The model also showed good calibration and clinical usage values. The NN model outperforms the XGBoost and RF models in predicting POST, with potential applications in the healthcare industry for reducing the incidence of this common postoperative complication.

Classification of Manners of the Prevocalic Alveolar Consonants in Machine Learning Using Dynamic Formant Transitions of Vowels in Korean Spontaneous Speech

The present study tried to classify the prevocalic manners of alveolar consonants using the formant transitions of the vowel in a Korean spontaneous speech corpus. Random forest and neural network models were trained and tested on selections of F1, F2, and F3 samples taken at the vowel onset and target of the vowel. It was found that prevocalic manners could not be manifested properly by the samples of any one or two formants, taken singly at the vowel onset or doubly at the vowel onset and target. Rather, both models trained on all the F1, F2, and F3 measurements taken doubly at the vowel onset and target, manifested the prevocalic manners robustly, though the random forest model outperformed the neural network model. The former model was further trained on additional predictors. Vowel identity facilitated model classification substan- tially more than F0, gender, speaking rate, and vowel duration. The contribution of the latter predictors was rather marginal.

A New Architecture Paradigm for Tool Wear Prediction during AISI 9840 Drilling Operation

In conventional machining processes based on the chip removal mechanism, the progressive wear of the tool determines a change of the geometric characteristics of the cutting edge. Tool wear is a complex phenomenon and related tool life depends on several factors, such as cutting parameters, lubrication, and tool-workpiece relative trajectories. Tool wear progression affects the quality of the machined parts, making the tool replacement necessary even before its breakage. Moreover, in industrial practice, tool replacement cannot depend on the subjectivity of the operator, thus, the definition of an optimized strategy for cutting tool wear monitoring before tool failure is mandatory. This work compares Random Forest and Neural Network models for predicting tool wear in drilling. For the development of predictive models, tool life tests were performed by drilling through holes on AISI 9840 steel parts, with a coated tungsten carbide drill of 8 mm of diameter and by using constant cutting parameters. Flank wear of the tool was monitored. A set of statistical features computed from the vibration signals, the acoustic emission signals, the power signals and the torque signals constitute the input of the algorithms. The classification accuracy was 88% and 91% for Neural Network and Random Forest models respectively. In order to correctly define the tool replacement policy during production, the developed Random Forest model has been implemented in an industry, through production management software, achieving promising results.

Leveraging available data for efficient exploration of materials space using Machine Learning: A case study for identifying rare earth-free permanent magnets

It is shown that even the relatively small amount of available materials data can be innovatively utilized to explore the materials space in order to identify materials with desired target properties. As an example of this, data from the novomag and Novamag databases are used to train random forest and neural network models which can predict thermodynamic stability, and magnetic properties of materials. Performance of these models are tested thoroughly and are found satisfactory. These models are subsequently used to interpolate within the above databases, and to extrapolate to parts of the materials composition and structure space not covered in these databases, to identify stable, magnetic materials that have large saturation magnetization and large easy-axis anisotropy. Screening 686 materials via the trained models, and subsequently performing first principles calculations, 21 new candidate materials for rare earth free permanent magnet are identified. Some of these materials have anisotropy constants as large as 5 and 6 MJ m−3, larger than that of the most widely used permanent magnet Nd2Fe14B. This simple approach can be used to screen materials with other functionalities in future.

Revolutionizing Solid-State NASICON Sodium Batteries: Enhanced Ionic Conductivity Estimation through Multivariate Experimental Parameters Leveraging Machine Learning.

Na superionic conductor (NASICON) materials hold promise as solid-state electrolytes due to their wide electrochemical stability and chemical durability. However, their limited ionic conductivity hinders their integration into sodium-ion batteries. The conventional approach to electrolyte design struggles with comprehending the intricate interactions of factors impacting conductivity, encompassing synthesis parameters, structural characteristics, and electronic descriptors. Herein, we explored the potential of machine learning in predicting ionic conductivity in NASICON. We compile a database of 211 datasets, covering 160 NASICON materials, and employ facile descriptors, including synthesis parameters, test conditions, molecular and structural attributes, and electronic properties. Random forest (RF) and neural network (NN) models were developed and optimized, with NN performing notably better, particularly with limited data (R2=0.820). Our analysis spotlighted the pivotal role of Na stoichiometric count in ionic conductivity. Furthermore, the NN algorithm highlighted the comparable significance of synthesis parameters to structural factors in determining conductivity. In contrast, the impact of electronegativity on doped elements appears less significant, underscoring the importance of dopant size and quantity. This work underscores the potential of machine learning in advancing NASICON electrolyte design for sodium-ion batteries, offering insights into conductivity drivers and a more efficient path to optimizing materials.

Advancing polytrauma care: developing and validating machine learning models for early mortality prediction

BackgroundRapid identification of high-risk polytrauma patients is crucial for early intervention and improved outcomes. This study aimed to develop and validate machine learning models for predicting 72 h mortality in adult polytrauma patients using readily available clinical parameters.MethodsA retrospective analysis was conducted on polytrauma patients from the Dryad database and our institution. Missing values pertinent to eligible individuals within the Dryad database were compensated for through the k-nearest neighbor algorithm, subsequently randomizing them into training and internal validation factions on a 7:3 ratio. The patients of our institution functioned as external validation cohorts. The predictive efficacy of random forest (RF), neural network, and XGBoost models was assessed through an exhaustive suite of performance indicators. The SHapley Additive exPlanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) methods were engaged to explain the supreme-performing model. Conclusively, restricted cubic spline analysis and multivariate logistic regression were employed as sensitivity analyses to verify the robustness of the findings.ResultsParameters including age, body mass index, Glasgow Coma Scale, Injury Severity Score, pH, base excess, and lactate emerged as pivotal predictors of 72 h mortality. The RF model exhibited unparalleled performance, boasting an area under the receiver operating characteristic curve (AUROC) of 0.87 (95% confidence interval [CI] 0.84–0.89), an area under the precision-recall curve (AUPRC) of 0.67 (95% CI 0.61–0.73), and an accuracy of 0.83 (95% CI 0.81–0.86) in the internal validation cohort, paralleled by an AUROC of 0.98 (95% CI 0.97–0.99), an AUPRC of 0.88 (95% CI 0.83–0.93), and an accuracy of 0.97 (95% CI 0.96–0.98) in the external validation cohort. It provided the highest net benefit in the decision curve analysis in relation to the other models. The outcomes of the sensitivity examinations were congruent with those inferred from SHAP and LIME.ConclusionsThe RF model exhibited the best performance in predicting 72 h mortality in adult polytrauma patients and has the potential to aid clinicians in identifying high-risk patients and guiding clinical decision-making.

Magnetic‐optical dual functional Janus particles for the detection of metal ions assisted by machine learning

AbstractFunctional polymer microspheres have broad application prospects in various fields, such as metal ion detection, adsorption, separation, and controlled drug release. However, integrating different functions in a single microsphere system is a significant challenge in this field. In this work, we prepared multicompartmental emulsion droplets utilizing microfluidic technology. Fe3O4 magnetic nanoparticles were added to one of the compartments of the emulsion droplets as functional particles, and Janus microspheres were obtained after curing. Fluorescent probes enter the two compartments of the Janus microspheres by diffusion. The fluorescence changes of the microspheres were observed in situ and captured through a fluorescence microscope. The images are processed by image recognition software and a Python program. The “fingerprint” of the detected metal ions is obtained by dimensionality reduction of the data through Principal Component Analysis. We employ different algorithms to build Machine Learning models for predicting the metal ion species and concentration. The variation of fluorescence intensity of the three fluorescent probes and the corresponding R, G, and B channel values and time are used as descriptors. The results show that the Random Forest, K‐neighborhood (KNN), and Neural Network models demonstrated a better predicted effect with a variance (R2) greater than 0.9 and a smaller root mean square error; among them, the KNN model predicted the most accurate results.

Incorporation of neighborhood information improves performance of SDB models

Mapping of water depth in nearshore waters is important for safe navigation and other human uses of coastal areas. Optical satellite imagery offers a cost-efficient source of data for this purpose. Several methods have been developed for satellite-derived bathymetry (SDB), but the details of their implementation, as well as their relative performance and its dependence on the data and environmental context, is unclear. We compared five SDB models, calibrated and tested against airborne lidar data from southern Corsica. We used a blocked strategy to split data into calibration, validation, and test data, because randomly splitting the data resulted in spatial autocorrelation between these data sets and underestimation of model prediction errors. In general, the non-parametric random forest and neural network models outperformed the parametric multi-band and band-ratio models. The convolutional neural network (CNN) model performed the best, especially in relatively deep (10–20 m) areas, and it could be near-optimally tuned with only ∼1000 data points. However, developing this CNN model was an iterative trial-and-error process that took several months. Using the random forest model, we demonstrated that using information from more than two bands in the visible and near-infrared spectrum contributed to improving model performance, as did incorporation of information from the local neighborhood. We suggest that the use of more than two bands, and the inclusion of spatial contextual information in addition to the values of the individual pixel, should become standard practice in SDB.

Differentiating brain states via multi-clip random fragment strategy-based interactive bidirectional recurrent neural network

EEG is widely adopted to study the brain and brain computer interface (BCI) for its non-invasiveness and low costs. Specifically EEG can be applied to differentiate brain states, which is important for better understanding the working mechanisms of the brain. Recurrent neural network (RNN)-based learning strategy has been widely utilized to differentiate brain states, because its optimization architectures improve the classification performance for differentiating brain states at the group level. However, present classification performance is still far from satisfactory. We have identified two major focal points for improvements: one is about organizing the input EEG signals, and the other is related to the design of the RNN architecture. To optimize the above-mentioned issues and achieve better brain state classification performance, we propose a novel multi-clip random fragment strategy-based interactive bidirectional recurrent neural network (McRFS-IBiRNN) model in this work. This model has two advantages over previous methods. First, the McRFS component is designed to re-organize the input EEG signals to make them more suitable for the RNN architecture. Second, the IBiRNN component is an innovative design to model the RNN layers with interaction connections to enhance the fusion of bidirectional features. By adopting the proposed model, promising brain states classification performances are obtained. For example, 96.97% and 99.34% of individual and group level four-category classification accuracies are successfully obtained on the EEG motor/imagery dataset, respectively. A 99.01% accuracy can be observed for four-category classification tasks with new subjects not seen before, which demonstrates the generalization of our proposed method. Compared with existing methods, our model outperforms them with superior results. Overall, the proposed McRFS-IBiRNN model demonstrates great superiority in differentiating brain states on EEG signals

Evaluation of the dataset quality in gamma passing rate predictions using machine learning methods.

Gamma passing rate (GPR) predictions using machine learning methods have been explored for treatment verification of radiotherapy plans. However, these methods presented datasets with unbalanced number of plans having different treatment conditions (heterogeneous datasets), such as anatomical sites or dose per fractions, leading to lower model interpretability and prediction performance. We investigated the impact of the dataset composition on GPR binary classification (pass/fail) using random forest (RF), XG-boost, and neural network (NN) models. 945 plans were used to create one reference dataset (randomly assembled) and 24 customized datasets that considered four heterogeneity factors independently (anatomical region, number of arcs, dose per fraction, and treatment unit). 309 predictor features were extracted and calculated from plan parameters, modulation complexity metrics, and radiomic analysis (leave-trajectory maps, 3D dose distributions, and portal dosimetry images). The models' performances were measured using the area under the curve from the receiver operating characteristic (ROC-AUC). Radiomics features for reference models increased ROC-AUC values up to 13%, 15%, and 5% for RF, XG-Boost, and NN, respectively. The datasets with higher heterogeneous conditions presented the lower ROC-AUC values (RF: 0.72 ± 0.11, XG-Boost: 0.67 ± 0.1, NN: 0.89 ± 0.05) compared to models with less heterogeneous treatment conditions (RF: 0.88 ± 0.06, XG-Boost: 0.89 ± 0.07, NN: 0.98 ± 0.01). The ten most important features for each heterogeneity dataset group demonstrated their correlation with the treatments' physical aspects and GPR prediction. Improvements in data generalization and model performances can be associated with datasets having similar treatment conditions. This analysis might be implemented to evaluate the dataset quality and model consistency of further ML applications in radiotherapy. Dataset heterogeneities decrease ML model performance and reliability.

Classification of Organic and Conventional Vegetables Using Machine Learning: A Case Study of Brinjal, Chili and Tomato

Growing organic food is becoming a challenging task with increasing demand. Food fraud activity has increased considerably with the increase in population growth. Consumers cannot visually distinguish between conventional and organically grown food products. Spectroscopic methodologies are presented to identify chemicals in food, thereby identifying organic and conventional food. Such spectroscopic techniques are laboratory-based, take more time to produce an outcome, and are costlier. Thus, this research designed a portable, low-cost multispectral sensor system to discriminate between organic and conventional vegetables. The designed multispectral sensor system uses a wavelength range (410 nm-940 nm) that includes three bands, namely visible (VIS), ultraviolet (UV) and near-infrared (NIR) spectra, to enhance the accuracy of detection. Tomato, brinjal and green chili samples are employed for the experiment. The organic and conventional discrimination problem is formulated as a classification problem and solved through random forest (RF) and neural network (NN) models, which achieve 92% and 89% accuracy, respectively. A two-stage enhancement mechanism is proposed to improve accuracy. In the first stage, the fuzzy logic mechanism generates additional feature sets. Ant colony optimization (ACO) algorithm-based parameter tuning and feature selection are employed in the second stage to enhance accuracy further. This two-stage improvement mechanism results in 100% accuracy in discriminating between organic and conventional vegetable samples. The detected adulterant is displayed on a web page through an IoT-developed application module to be accessed from anywhere.

Study on Influencing Factors of Ethanol Coupling to Prepare C4 Olefins

In this paper, we study the data analysis and calculation of the factors influencing the preparation of C4 olefins by ethanol coupling. Specifically, by combining each catalyst, the relationship between temperature and ethanol conversion and C4 olefin selectivity was analyzed by regression analysis and correlation analysis, respectively. In addition, the grey relational degree analysis was used to solve the grey relational degree between the influencing factors of the experiment. The nonlinear correspondence between ethanol conversion and C4 olefin selectivity with different catalyst combinations and temperatures was evaluated by means of random forest model and BP neural network model. Finally, we build a support vector machine regression (SVMR) model and optimize the cost and gamma parameters of this model using a genetic algorithm.

A New Approach to Machine Learning Model Development for Prediction of Concrete Fatigue Life under Uniaxial Compression

The goal of this work is to show how machine learning models, such as the random forest, neural network, gradient boosting, and AdaBoost models, can be used to forecast the fatigue life (N) of plain concrete under uniaxial compression. Here, we developed our final machine learning model by generating the following three data files from the original data used in the work of Zhang et al.: (a) grouped data with the same input variable value and different output variable logN value, (b) data excluding outliers selected by three or more outlier detection methods; (c) average data excluding outliers, created by averaging the grouped data after excluding outliers from among the grouped data. Excluding the sustained strength of the concrete variable, originally treated as the seventh input variable in the work of Zhang et al., resulted in improving the determination coefficient (R2) values. Moreover, the gradient boosting model showed a high R2 value at 0.753, indicating a high accuracy in predicting outcomes. Further analysis using data excluding outliers shows that the R2 value increased to 0.803. Moreover, the average data excluding outliers provided the best R2 value at 0.915. Finally, a permutation feature importance (PFI) analysis was carried out to determine the strength of the relationship between the feature and the target value for the gradient boosting model. The analysis results showed that the maximum stress level (Smax) and loading frequency (f) were the most significant input variables, followed by compressive strength (f′c) and maximum to minimum stress ratio (R). Shape and height to width ratio (h/w) were the features with a non-significant influence on the model. This trend was previously confirmed by a Pearson and Spearman correlation analysis.

A Neural Network Model for the Relationship between Hotel Marketing Strategies and Performance Based on Nonlinear Random Matrix Theory

The homogenization of hotel competition has become a key problem faced by the hotel enterprises. According to the marketing theory, one perfect marketing strategy will improve hotel performance. Based on nonlinear random matrix and neural network model, this paper empirically analyzes the relationship between different marketing strategies and hotel performance from the aspects of hotel demand collection, strategy support, performance matching, and strategy regulation. The results show that the hotel performance based on product marketing strategies is the best, which has improving hotel performance about 5%. In contrast, the hotel performance based on people-oriented marketing strategies is the worst. Therefore, we believe that the hotel enterprises should pay attention to the adjustment of product structure and the role of people in hotel operation and management. Adjusting product structure, strengthening hotel information construction, and paying attention to talent cultivation play an important role in improving hotel performance, in order to provide useful reference for the development of other hotels.

Evaluation of Physical Education Teaching Quality Based on the Random Multivariate Matrix Convolution Neural Network Model

With the non-stop merchandising and improvement of bodily training instructing reform, the institution of a couple of getting to know the assessment and evaluation mannequin of bodily training different educating impact can stimulate the enthusiasm and activity in sports activities and learning, which is useful to domesticate students’ cognizance of lifelong bodily exercise. In the educating process, through the integration technique of varied instructing methods, this paper explores the issues present in the method of bodily training reform, places ahead a diverse bodily training instructing impact assessment mannequin that meets the wants of customized intelligence education and construction, and makes use of the technique of empirical evaluation to affirm the comparison model. In order to reflect the effect of physical training effectively and accurately, a convolution neural neighbourhood model based totally completely on the random multivariate matrix is proposed to reflect on consideration on the excellent of bodily education. The overall performance of the prediction accuracy assessment mannequin is evaluated via the simulation test and in contrast with the standard approach model. The experimental information is that the average assessment accuracy of the single convolution neural community mannequin is 82.15%, whilst the common comparison accuracy of the convolution neural community mannequin based totally on the random multivariate matrix is 97.58%, and the prediction accuracy is increased by means of 15.43%. The common prediction error of the single convolution neural community mannequin is 0.97, and the common error is 0.91, the common error is decreased by using 0.05. It shows that the random multivariate matrix convolution neural neighbourhood model can efficaciously realize the evaluation of instructing quality.