Neural Network Algorithm Research Articles

Research on the Recognition of Piano-Playing Notes by a Music Transcription Algorithm

With the deepening research on music works, music transcription algorithms have been increasingly studied. This study examined the recognition of piano-playing notes using a music-transcription algorithm. First, the characteristics of MelSpec, LogSpec, and the constant Q-transform (CQT) are briefly introduced. Then, a convolutional recurrent neural network (CRNN) transcription algorithm, which includes four convolutional blocks and one bidirectional long short-term memory (BiLSTM) structure, was designed. The recognition performance of this method was analyzed using the MAPS dataset. LogSpec was found to have the best recognition performance for piano-playing notes when used as an input feature. In the CRNN structure, the recognition performance for piano-playing notes was the best when four convolutional blocks were used. Compared with the convolutional neural network (CNN), BiLSTM, and CNN-hidden Markov model algorithms, the F1-values of the CRNN algorithm were 84.9%, 92.24%, and 79.27% for frames, notes, and offsets, respectively, achieving the best recognition results. The results verify that the CRNN transcription algorithm is effective for the recognition of piano-playing notes and can be applied in practice.

Research on Cross-Border e-Commerce Supply Chain Prediction and Optimization Model Based on Convolutional Neural Network Algorithm

Enhancing the precision of supply chain management and reducing operational costs are crucial for the development of the cross-border e-commerce market. However, existing research often overlooks the demand uncertainty caused by seasonal variations and the challenges of handling returns in logistics. Therefore, this paper proposes a SARIMA-CNN-BiLSTM prediction model that effectively captures both the seasonal and nonlinear characteristics of cross-border e-commerce supply chains. Additionally, by incorporating the returns process, a supply chain distribution optimization model is developed with the objective of minimizing total operational costs. The model is solved using an improved whale optimization algorithm. In validation with real-world data, the SARIMA-CNN-BiLSTM model achieved a mean absolute percentage error reduction of 6.479 and 7.703 compared to convolutional neural network (CNN) and BiLSTM models, respectively. Moreover, the chosen optimization algorithm reduced the cost by 231,310 CNY, 62,564 CNY, and 131,632 CNY compared to the whale optimization algorithm, genetic algorithm, and particle swarm optimization, respectively. The proposed approach provides robust support for cross-border e-commerce enterprises in reducing costs and enhancing efficiency in their operations.

Improving Multiclass Rainfall Prediction with Multilayer Perceptron and SMOTE: Addressing Class Imbalance Challenges

Rainfall is a key climate element that affects weather patterns and human activities, especially in agriculture and daily life. Therefore, accurately classifying rainfall is crucial for predicting future rainfall amounts. This study uses the Multilayer Perceptron (MLP) classification method, a neural network algorithm, to classify rainfall. The dataset, sourced from the BMKG website, has a class imbalance, requiring using the SMOTE (Synthetic Minority Over-sampling Technique) technique. The research compares the performance of MLP with and without SMOTE. The results show that the best model was achieved with SMOTE. MLP without SMOTE achieved an accuracy of 75%, sensitivity of 40.34%, specificity of 86.15%, and an AUC of 63.25%. In comparison, MLP with SMOTE achieved an accuracy of 71.27%, sensitivity of 71.14%, specificity of 90.30%, and an AUC of 80.72%. Although accuracy decreased, the overall evaluation, particularly the AUC, improved significantly. Therefore, the SMOTE technique effectively addresses the class imbalance issue in rainfall classification.

Inference of Behavioral Decision-Making Using AI and Statistical Data Analysis in the Big Data Environment

Big data are of essence to enhance the accuracy of making decisions and in improving the ability to forecast. However, working with large and complicated datasets, the traditional methods of analyzing data tend to be grossly inadequate. It is at this point, amongst others, that artificial intelligence has now presented a feasible solution to such limitations, especially with the model called machine learning. Based on that, this research will look into the aspect of integration between artificial intelligence and statistical methods of analysis in inferring behavioral decisions from big data. This work considers a number of datasets related to marketing, health care, and finance, comparing the efficiency of the application of a range of artificial intelligence (AI) models, especially random forests, Long Short Term Memory (LSTM), and convolutional neural network (CNN) algorithms, against the classical statistical ones. Standard performance evaluation indicators apply: accuracy, precision, recall rate, F1 score are applied in the model. Results have shown that, though model interpretability and overfitting are challenges, the predictive accuracy of artificial intelligence models is somewhat better in comparison with conventional statistical methods. The present study underlines the transformative potential of AI in changing decision-making across many industries but also highlights key areas for further improvement on behalf of real-time processing abilities and ethical deployment consideration.

Exercise training load prediction based on improved genetic algorithm

With the advancement of digital information age, all kinds of sports also widely used digital mining technology, the technology can promote athletes’ professional level and physical quality, is in the field of sports improve athletes and coaches more effective way, but the bicycle sports for digital mining and research is still in the early state. Cyclists load ability in training is mainly influenced by the body function, the relationship between the two is complicated, this paper is the training load ability and physical function as a research object, explore the specific connection between the two aspects of cyclists, by constructing the cyclist training load prediction model, to apply in the “bicycle team training analysis system”. Strive to promote the overall development of cycling, to provide a basis for the scientific training of talents and training. The cyclist training load prediction model created in this paper can provide a scientific design basis for the training of athletes. When analyzing the body function, this paper mainly analyzes the influence of athletes on conforming sports through 25 aspects such as maximum oxygen intake, functional threshold power and blood oxygen saturation. Because the factors of athletes and the predicted results show non-linear correlation, this paper selects the BP neural network algorithm in the model algorithm, and the adaptive genetic algorithm of the selection operator is adjusted and optimized to clarify the initial weights and thresholds of BP neural network. Finally, the practice has proved that the improved algorithm has a more prominent global optimization ability, and the accuracy of the model has reached 93.28%.

Spatial Coupling Relationship Between Water Area and Water Level of Dongting Lake Based on Multiple Temporal Remote Sensing Images Data at Its Several Hydrological Stations

It is very well-known that the reliable coupling relationship between water area and water level is very important in analyzing the risks of floods and droughts for big lakes, such as Dongting Lake, especially when remote sensing images are absent and in situ measurements cannot be carried out. To obtain this relationship, two types of mathematical models—polynomial regression (PR) based on the least square algorithm and machine learning regression (MLR) based on the BP (Backpropagation) neural network algorithm—are constructed using the water area data extracted from multiple temporal remote sensing images and water levels recorded at several representative hydrological stations for nearly 30 years. In this study, Dongting Lake is divided into three parts: East Dongting Lake (EDL), South Dongting Lake (SDL), and West Dongting Lake (WDL). This is because water slope exists on its surface, which is formed by several inflow rivers and the high and low terrain. To calculate the total water area of this lake, two ways are put forward by choosing the water levels: from EDL, SDL, and WDL in their turn; or from all three simultaneously. In other words, three univariate and one multivariate regression. For PR, there are perfect coefficients of determination (most nearly 0.95, the smallest being 0.76), which is in line with regression test relative errors (between 0.27% and 6.7%). For MLR, which was initially applied to this problem, the best node number (10 for the first way, 8 for the second way) in the hidden layer of the neural network is adaptively chosen, with coefficients of determination (similar to PR), together with training and testing error performances (between 1% and 10%). These results confirm the validity and reliability of them. The regression and prediction results on the two models are better than the documented way (only focus on the water level of EDL). These results can provide some references for researchers and decision makers in studying similar big Lakes.

Comparing prediction accuracy for 30-day readmission following primary total knee arthroplasty: the ACS-NSQIP risk calculator versus a novel artificial neural network model

BackgroundUnplanned readmission, a measure of surgical quality, occurs after 4.8% of primary total knee arthroplasties (TKA). Although the prediction of individualized readmission risk may inform appropriate preoperative interventions, current predictive models, such as the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) surgical risk calculator (SRC), have limited utility. This study aims to compare the predictive accuracy of the SRC with a novel artificial neural network (ANN) algorithm for 30-day readmission after primary TKA, using the same set of clinical variables from a large national database.MethodsPatients undergoing primary TKA between 2013 and 2020 were identified from the ACS-NSQIP database and randomly stratified into training and validation cohorts. The ANN was developed using data from the training cohort with fivefold cross-validation performed five times. ANN and SRC performance were subsequently evaluated in the distinct validation cohort, and predictive performance was compared on the basis of discrimination, calibration, accuracy, and clinical utility.ResultsThe overall cohort consisted of 365,394 patients (trainingN = 362,559; validationN = 2835), with 11,392 (3.1%) readmitted within 30 days. While the ANN demonstrated good discrimination and calibration (area under the curve (AUC)ANN = 0.72, slope = 1.32, intercept = −0.09) in the validation cohort, the SRC demonstrated poor discrimination (AUCSRC = 0.55) and underestimated readmission risk (slope = −0.21, intercept = 0.04). Although both models possessed similar accuracy (Brier score: ANN = 0.03; SRC = 0.02), only the ANN demonstrated a higher net benefit than intervening in all or no patients on the decision curve analysis. The strongest predictors of readmission were body mass index (> 33.5 kg/m2), age (> 69 years), and male sex.ConclusionsThis study demonstrates the superior predictive ability and potential clinical utility of the ANN over the conventional SRC when constrained to the same variables. By identifying the most important predictors of readmission following TKA, our findings may assist in the development of novel clinical decision support tools, potentially improving preoperative counseling and postoperative monitoring practices in at-risk patients.

Leveraging natural language processing for automated assessment and feedback production in virtual education settings

With the rapid development of online teaching, the traditional teacher face-to-face evaluation and feedback model in online teaching faces problems such as high labor costs, low efficiency, and insufficient personalization. The application of natural language processing technology in the field of education has become a trend. Based on this choice, this article applied natural language processing to achieve automated evaluation and personalized feedback of online teaching processes. This article chose to use recurrent neural network (RNN) algorithm for text generation, and adopted Seq2Seq-based (Sequence-to-Sequence) recurrent neural network framework for automatic text feedback generation. Then, a group of student performance data was collected and divided into a control group and an experimental group, and comparisons were made between manual feedback and automatically generated feedback, respectively. Finally, the effectiveness and impact of automatic evaluation and feedback generation were evaluated by comparing the performance data and analysis results of two groups of students. The average accuracy of the automatic evaluation system designed in this article was 95.6%, which was higher than the 84.5% of manual evaluation. The accuracy of the automatic evaluation system designed in this article was higher, and the time spent on manual evaluation and automatic evaluation was compared. The average time spent on manual evaluation was 62.2 seconds, while automatic evaluation shortened the time to 36.98 seconds. For the feedback results, the graphical curves of the manually generated and automatically generated feedback results in terms of feedback on student grades were basically consistent, which can fully prove that the automatically generated feedback results have high accuracy. The remarkable efficacy of natural language processing (NLP) in online education is vividly illustrated by a recent study, where an NLP model notched an impressive average ROUGE score of 93.8 points. This achievement outstrips the performance of three other algorithmic models in the realm of automatic assessment and feedback generation. Such comparative experiments underscore the transformative potential of NLP technology, significantly boosting teaching efficiency and enriching the learning experience by adeptly tackling the perennial challenges of evaluation and feedback in online education settings.

Application research of convolutional neural network and its optimization in lightning electric field waveform recognition

Quickly identifying and classifying lightning waveforms is the foundation of lightning forecasting and early warning. In this paper, based on the electric field observation of the Beijing lightning location website of the Institute of Atmospheric Physics, Chinese Academy of Sciences, a recognition and classification method of pulse signal waveform based on Convolutional Neural Network(CNN) algorithm is designed and implemented. The CNN network model and its parameters were optimized from three aspects: dataset, model parameters, and network structure, achieving a recognition rate of over 90%. The effects of various optimization terms and their different optimization orders on the training time of the model were studied. The results indicate that the CNN algorithm is suitable for the classification and recognition of lightning electric field (LEF) waveforms. Optimization can significantly improve recognition rate. The optimization method of fitting idealized waveforms can reduce noise in the dataset and significantly improve recognition rate, indicating that noise has a significant impact on waveform recognition. Therefore, it is necessary to perform noise preprocessing before recognition. The optimization has a huge impact on training efficiency, increasing training time by about 51% after optimization, but the influence of optimization order on it can be ignored.

Evaluation and Analysis of Particle Oxidation of HVOF Thermal Spraying Based on GA-BP Neural Network Algorithm

Evaluation and Analysis of Particle Oxidation of HVOF Thermal Spraying Based on GA-BP Neural Network Algorithm

Analysis and Research on the Intelligent Control System of Comprehensive Traffic Signals

The existing urban traffic signal control system exhibits certain inefficiencies, including prolonged waiting times and frequent traffic congestion. Research indicates that the integration of neural networks can enhance the effectiveness of intelligent traffic control systems, potentially improving overall traffic flow and reducing delays. Neural network is a highly parallel, nonlinear and highly redundant system. Meanwhile, the self-learning and self-organizing capabilities of neural network have unique advantages for those systems that cannot establish accurate mathematical models. According to the characteristics of complex, variable and random traffic conditions at urban intersections, this paper analyzes the traffic signal control by using neural network control method. This paper introduces a design method of intelligent traffic light based on dynamic time adjustment of traffic flow. By detecting the traffic flow on site, and then allocating the green time of each lane according to the neural network algorithm, the traffic flow dynamic adjustment can be realized, and the traffic capacity of the intersection can be effectively improved. This kind of intelligent control traffic lights can effectively control traffic, and has obvious effects on dreading traffic flow, improving road capacity and reducing traffic accidents.

Construction of intelligent coal blending system based on BP neural network and genetic algorithm

The quality of coke is crucial to the steel-making process, and advanced coal blending technology can effectively improve the stability of coke and reduce production costs. However, some of the current coal blending systems lack expertise and are out of touch with the actual situation of the factory, unable to meet production requirements. In the research process, this paper combined the conventional indicators measured in Chinese factories with actual data, and established a reliable database. In the back propagation neural network modeling process, several parameters were optimised, significantly improving the prediction performance compared to traditional regression methods. An evolutionary algorithm with a penalty matrix is used to optimise the coal blending ratio, resulting in a reduction in the ash and sulphur content of the coke, as well as a cost reduction of approximately 44 RMB per tonne. Based on the initial research results, an intelligent coal blending system was developed, which has been successfully applied in the production process of a coke plant, with simple operation and stable operation, providing important reference for the development of coal blending systems in other coke plants.

Prediction of Postpartum Depression With Dataset Using Hybrid Data Mining Classification Technique

Postpartum Depression is a condition or a state which usually affects the woman immediately after child birth. The birth of a baby not only brings delighted emotions such as excitement, but also fear and anxiety which may sometimes lead to depression. It is a period of physical, emotional and behavioral changes that happen in some woman immediately after the delivery. Apart from the chemical changes, there are many factors which affect a woman during and after pregnancy period. If PPD is not identified and treated at the earlier stages, it may lead to serious issues for mother and child. It is therefore of vital importance to sift through the woman at any early stage to prevent any consequences. The objective of this study is to find out the presence of PPD without getting worse. Data mining plays an important role in the health care industry with successful outcome. It helps to find out hidden patterns, trends and anomalies from large dataset to make the predictions. The proposed system is a combined classification technique for the prediction of postpartum depression that uses Support vector machine, Artificial Neural Network and Hybrid classifier algorithm to produce the best result.

Reliability-based reinforcement learning driven maintenance policy optimization

The study presents a reinforcement learning (RL) method called the Q-learning algorithm to determine the best maintenance policy for equipment. This involves an artificial intelligence agent making decisions and an environment representing the equipment. The agent creates maintenance policies and takes actions, while the environment determines state transitions and rewards based on the actions chosen. The optimization of the maintenance policy starts with predicting equipment reliability using sensor data. This prediction method combines Back Propagation Neural Network (BPNN) algorithms with the Boxing Match Algorithm (BMA), an evolutionary meta-heuristic algorithm. A novel weight update strategy enhances the performance of artificial neural networks in reliability prediction. This integrated model, BMA-BPNN, aims to improve the accuracy of forecasted reliability. The study involves forecasting equipment reliability to determine critical levels, incorporating cost and risk considerations into decision-making for optimizing maintenance policies. As a result, the Q-learning algorithm is used to identify the best maintenance actions based on equipment reliability. Implementing an automated maintenance system that considers equipment reliability and costs can help reduce accidents resulting from maintenance program deficiencies. This study thus contributes to the field by providing an efficient approach to equipment maintenance.

Improving Recall Accuracy in Sparse Associative Memories That Use Neurogenesis.

The creation of future low-power neuromorphic solutions requires specialist spiking neural network (SNN) algorithms that are optimized for neuromorphic settings. One such algorithmic challenge is the ability to recall learned patterns from their noisy variants. Solutions to this problem may be required to memorize vast numbers of patterns based on limited training data and subsequently recall the patterns in the presence of noise. To solve this problem, previous work has explored sparse associative memory (SAM)-associative memory neural models that exploit the principle of sparse neural coding observed in the brain. Research into a subcategory of SAM has been inspired by the biological process of adult neurogenesis, whereby new neurons are generated to facilitate adaptive and effective lifelong learning. Although these neurogenesis models have been demonstrated in previous research, they have limitations in terms of recall memory capacity and robustness to noise. In this letter, we provide a unifying framework for characterizing a type of SAM network that has been pretrained using a learning strategy that incorporated a simple neurogenesis model. Using this characterization, we formally define network topology and threshold optimization methods to empirically demonstrate greater than 10$^{{4}}$ times improvement in memory capacity compared to previous work. We show that these optimizations can facilitate the development of networks that have reduced interneuron connectivity while maintaining high recall efficacy. This paves the way for ongoing research into fast, effective, low-power realizations of associative memory on neuromorphic platforms.

Multiparametric MRI for Assessment of the Biological Invasiveness and Prognosis of Pancreatic Ductal Adenocarcinoma in the Era of Artificial Intelligence.

Pancreatic ductal adenocarcinoma (PDAC) is the deadliest malignant tumor, with a grim 5-year overall survival rate of about 12%. As its incidence and mortality rates rise, it is likely to become the second-leading cause of cancer-related death. The radiological assessment determined the stage and management of PDAC. However, it is a highly heterogeneous disease with the complexity of the tumor microenvironment, and it is challenging to adequately reflect the biological aggressiveness and prognosis accurately through morphological evaluation alone. With the dramatic development of artificial intelligence (AI), multiparametric magnetic resonance imaging (mpMRI) using specific contrast media and special techniques can provide morphological and functional information with high image quality and become a powerful tool in quantifying intratumor characteristics. Besides, AI has been widespread in the field of medical imaging analysis. Radiomics is the high-throughput mining of quantitative image features from medical imaging that enables data to be extracted and applied for better decision support. Deep learning is a subset of artificial neural network algorithms that can automatically learn feature representations from data. AI-enabled imaging biomarkers of mpMRI have enormous promise to bridge the gap between medical imaging and personalized medicine and demonstrate huge advantages in predicting biological characteristics and the prognosis of PDAC. However, current AI-based models of PDAC operate mainly in the realm of a single modality with a relatively small sample size, and the technical reproducibility and biological interpretation present a barrage of new potential challenges. In the future, the integration of multi-omics data, such as radiomics and genomics, alongside the establishment of standardized analytical frameworks will provide opportunities to increase the robustness and interpretability of AI-enabled image biomarkers and bring these biomarkers closer to clinical practice. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 4.

Integrating machine learning and nano-QSAR models to predict the oxidative stress potential caused by single and mixed carbon nanomaterials in algal cells.

In silico methods are increasingly important in predicting the ecotoxicity of engineered nanomaterials (ENMs), encompassing both individual and mixture toxicity predictions. It is widely recognized that ENMs trigger oxidative stress effects by generating intracellular reactive oxygen species (ROS), serving as a key mechanism in their cytotoxicity studies. However, existing in silico methods still face significant challenges in predicting the oxidative stress effects induced by ENMs. Herein, we utilized laboratory-derived toxicity data and machine learning methods to develop quantitative nanostructure-activity relationship (nano-QSAR) classification and regression models, aiming to predict the oxidative stress effects of five carbon nanomaterials (fullerene, graphene, graphene oxide, single-walled carbon nanotubes, and multi-walled carbon nanotubes) and their binary mixtures on Scenedesmus obliquus cells. We constructed five nano-QSAR classification models by combining zeta potential (ζP) with the C4.5 decision tree, support vector machine, artificial neural network, naive Bayes, and K-nearest neighbor algorithms. Moreover, we constructed three classification models by integrating the features including ζP, hydrodynamic diameter (DH), and specific surface area (SSA) with the logistic regression, random forest, and Adaboost algorithms. The Accuracy, Recall, Precision and harmonic mean of Precision and Recall (F1-score) values of these models were all higher than 0.600, indicating an excellent performance in distinguishing whether CNMs have the potential to generate ROS. In addition, using the ζP, DH, and SSA descriptors, we combined decision tree regression, random forest regression, gradient boosting, and the Adaboost algorithm, and successfully constructed four nano-QSAR regression models with applicable application domains (all training and testing data points lie within 95% confidence intervals), goodness-of-fit (Rtrain2 ≥ 0.850), and robustness (cross-validation R2 ≥ 0.650) as well as predictive power (Rtest2 ≥ 0.610). The method developed would establish a fundamental basis for more precise evaluations of ecological risks posed by these materials from a mechanistic standpoint.

Arrears behavior prediction of power users based on BP neural network and multi-scale feature learning: a refined risk assessment framework

This study aims to develop an efficient model to predict the arrears behavior of electricity users by integrating multi-scale feature learning with a backpropagation (BP) neural network. The goal is to provide accurate early warning systems and enhanced risk management tools for power companies. The BP neural network algorithm adjusts weights to minimize prediction errors, while multi-scale feature learning captures the diversity and regularity of user behavior by extracting data from various time dimensions, such as daily, weekly, and monthly intervals. First, electricity usage and weather data from the UMass Smart Dataset are preprocessed, including steps such as data cleaning, standardization, and normalization. Next, features are extracted across three time scales—daily, weekly, and monthly. These features are then input into the BP neural network model using the multi-scale feature learning method. A hierarchical neural network structure is designed to address the characteristics of different scales in distinct layers. Key model parameters are optimized, and a sensitivity analysis is conducted. The experimental results demonstrate that the BP neural network model incorporating multi-scale features outperforms traditional BP neural network models and other control models in several evaluation metrics. Specifically, the Gini coefficient is 0.55, the Kolmogorov-Smirnov statistic is 0.60, the Matthews correlation coefficient is 0.45, and specificity is 0.82. These results indicate that the proposed method offers significant improvements in capturing user behavior patterns and enhancing prediction accuracy. The study concludes that the effective fusion of multi-scale features not only enhances the model’s prediction performance but also strengthens its generalization ability. This method provides an advanced risk management tool for power companies, helping to increase the operational efficiency of smart grids and encouraging further research toward greater intelligence in the field.

Evaluate Lending Club Loan Status by Machine Learning

To reduce the information asymmetry between investors and lenders and enable P2P lending platforms to develop more accurate lending standards, this study uses Lending Club data to build a credit scoring model based on machine learning (ML) and artificial neural network (ANN) algorithms. The data is first cleaned and preprocessed through feature engineering techniques, and then trained using XGBoost to evaluate feature importance. To address the class imbalance problem, undersampling is used to divide the dataset into training and test sets. The study also uses grid search and cross-validation methods to determine the best hyperparameters for four algorithms (logistic regression (LR), random forest (RF), lightweight gradient boosting machine (LightGBM), and XGBoost). Finally, this paper compares and analyzes the performance of different models and points out the practicality of different models. The results of this process provide insights into the most important features for predicting creditworthiness and identify the most effective model for improving lending decisions on P2P platforms.

A Neural Network approach to optimize flow rate for Micro-Bubble Drag Reduction

To reduce the resistance of ships, many scholars have focused on micro-bubble drag reduction (MBDR) technology. However, The implementation of MBDR on ships is challenging due to the numerous factors and complex coupling relationships involved in the gas–liquid two-phase flow at the ship’s bottom. This paper aims to apply neural network algorithm on MBDR technology to real ships to achieve energy savings and emission reductions. A test platform was constructed, and the Micro-Bubble Flow Rate of the test model at the lowest drag resistance was collected to form a sample database. Using the database as a sample, the mathematical model is trained by the neural network algorithm, and optimized by the genetic algorithm through a number of iterations, the Optimal Micro-Bubbles Flow Rate (OMFR) is predicted for each status. This paper is of great significance to the application of MBDR technology on real ships. Using neural network algorithm to form the mathematics model from the data samples of unconstrained self-propulsion tests. The OMFR obtained through the iterative process of the genetic algorithm, can reduce the ship resistance, thereby achieving energy savings and emission reductions.