Network Machine Research Articles

Effect of Machine Learning Algorithms on Prediction of In-Cylinder Combustion Pressure of Ammonia–Oxygen in a Constant-Volume Combustion Chamber

Road vehicles, particularly cars, are one of the primary sources of CO2 emissions in the transport sector. Shifting to unconventional energy sources such as solar and wind power may reduce their carbon footprints considerably. Consequently, using ammonia as a fuel due to its potential benefits, such as its high energy density, being a carbon-free fuel, and its versatility during storage and transportation, has now grabbed the attention of researchers. However, its slow combustion speed, larger combustion chamber requirements, ignition difficulties, and limited combustion stability are still major challenges. Therefore, authors tried to analyze the combustion pressure of ammonia in a constant-volume combustion chamber across different equivalence ratios by adopting a machine learning approach. While conducting the analysis, the experimental values were assessed and subsequently utilized to predict the induced combustion pressure in a constant-volume combustion chamber across various equivalence ratios. In this research, a two-step prediction process was employed. In the initial step, the Random Forest algorithm was applied to assess the combustion pressure. Subsequently, in the second step, artificial neural network machine learning algorithms were employed to pinpoint the most effective algorithm with a lower root-mean-square error and R2. Finally, Linear Regression illustrated the lowest error in both steps with a value of 1.0, followed by Random Forest.

Abstract TP14: SocialBit: A Novel Smartwatch Sensor to Detect Social Interaction Frequency in Stroke Survivors

Introduction: Stroke survivors experience high levels of social isolation which hinder their rehabilitation and well-being. A gap in the field that impedes intervention development is a lack of real-time social interaction measurement tailored for this population. In response, we developed SocialBit, a smartwatch-based machine learning algorithm to discreetly and passively detect social interactions through privacy-preserving acoustic analysis. In this observational study, we estimated the accuracy of SocialBit to detect social interactions compared to human observers in stroke survivors. Methods: SocialBit was built as a neural network machine learning algorithm using features from YAMNet and a Transformer classifier. It was then trained and tested in stroke survivors for up to 8 days in hospital. Independently, human observers tallied the presence or absence of social interactions every minute to establish the ground truth. SocialBit performance of detecting social interactions was compared against the ground truth. Results: We analyzed 1137.7 hours of ambient audio data in 90 patients within 14 days of stroke onset. 16% of patients had aphasia, 25% reported symptoms of clinical depression, and 28% reported moderate to severe loneliness. SocialBit had an accuracy of 0.83 (SD = 0.2), sensitivity of 0.86 (SD = 0.2), specificity of 0.81 (SD = 0.02), and an area under the curve (AUC) of 0.90 (SD = 0.2) (Figure 1). Conclusion: SocialBit accurately detected the frequency of social interactions in stroke survivors in a hospital setting. These results will drive future planned studies to examine SocialBit's performance in patients’ home communities. SocialBit shows the potential to be the first real-time social sensor to quantify social interactions and social isolation in stroke survivors.

Hydro-chemical based assessment of groundwater vulnerability in the Holocene multi-aquifers of Ganges delta

Determining the degree of high groundwater arsenic (As) and fluoride (F−) risk is crucial for successful groundwater management and protection of public health, as elevated contamination in groundwater poses a risk to the environment and human health. It is a fact that several non-point sources of pollutants contaminate the groundwater of the multi-aquifers of the Ganges delta. This study used logistic regression (LR), random forest (RF) and artificial neural network (ANN) machine learning algorithm to evaluate groundwater vulnerability in the Holocene multi-layered aquifers of Ganges delta, which is part of the Indo-Bangladesh region. Fifteen hydro-chemical data were used for modelling purposes and sophisticated statistical tests were carried out to check the dataset regarding their dependent relationships. ANN performed best with an AUC of 0.902 in the validation dataset and prepared a groundwater vulnerability map accordingly. The spatial distribution of the vulnerability map indicates that eastern and some isolated south-eastern and central middle portions are very vulnerable in terms of As and F− concentration. The overall prediction demonstrates that 29% of the areal coverage of the Ganges delta is very vulnerable to As and F− contents. Finally, this study discusses major contamination categories, rising security issues, and problems related to groundwater quality globally. Henceforth, groundwater quality monitoring must be significantly improved to successfully detect and reduce hazards to groundwater from past, present, and future contamination.

Balancing urban energy considering economic growth and environmental sustainability through integration of renewable energy

This paper delves into the critical intersection of urban energy management, economic growth, and environmental sustainability through the integration of renewable energy sources in smart urban homes. The escalating demand for power has created an energy scarcity, necessitating innovative solutions to balance supply and consumption. In this context, the transition from traditional electric grids to smart grids (SG) is gaining prominence. To harness the potential of smart urban homes, particularly in the context of renewable energy, solar panels on rooftops have become indispensable. Smart grid technologies enable users to engage in demand response strategies, offering incentives and financial benefits for flexible device usage. This study introduces a novel optimization algorithm, the Polar Bear Optimization Algorithm (PBOA), designed to schedule domestic device usage patterns efficiently. To evaluate the effectiveness of PBOA, this paper conducts comparative analyses with other optimization algorithms such as Differential Evolution (DE) and Particle Swarm Optimization Algorithm (PSOA). Energy cost metrics, including Critical Peak Price (CPP) and Real-Time Price (RTP), are employed to assess cost reduction strategies. The primary focus of this study is to minimize energy costs and reduce the Peak to Average Ratio (PAR), thereby promoting economic growth and environmental sustainability in smart urban homes. It should be noted that neural network (NN) machine learning technique has been used for forecasting the weather and renewable energies output power.

WDM C-band four channel demultiplexer using cascaded multimode interference on SiN strip waveguide structure

Back reflection challenges significantly constrain the efficiency of optical communication networks employing dense wavelength division multiplexing (DWDM) technology, particularly those based on Silicon (Si) Multimode Interference (MMI) waveguides. To mitigate this limitation, we present a novel 1×4 optical demultiplexer design using MMI within a Silicon-Nitride (SiN) strip waveguide configuration, operating within the C-band spectrum. Our design was optimized using AI-enhanced RSoft-CAD simulations that combined the Beam Propagation Method (BPM) and Finite-Difference Time-Domain (FDTD) techniques, integrated with convolutional neural network (CNN) machine learning algorithms. The simulation results demonstrate that the proposed device efficiently transmits four channels with 10 nm spacing in the C-band, showing low power loss ranging from 1.98-2.35 dB, a broad bandwidth of 7.68-8.08 nm, and excellent crosstalk suppression between 20.8-23.4 dB. Leveraging the low refractive index of SiN, we achieve ultra-low back reflection of 40.58 dB without requiring specialized angled MMI designs, which are often necessary in Si-based MMI technology. Consequently, this SiN-based MMI demultiplexer provides an effective solution for DWDM systems, enabling high data transmission rates with minimal back reflection in optical communication networks.

Utilizing machine learning to enhance performance of thin-film solar cells based on Sb2(S x Se1-x )3: investigating the influence of material properties.

Antimony chalcogenides (Sb2(S x Se1-x )3) have drawn attention as a potential semiconducting substance for heterojunction photovoltaic (PV) devices due to the remarkable optoelectronic properties and wide range of bandgaps spanning from 1.1 to 1.7 eV. In this investigation, SCAPS-1D simulation software is employed to design an earth abundant, non-toxic, and cost-effective antimony sulfide-selenide (Sb2(S,Se)3)-based thin-film solar cell (TFSC), where tungsten disulfide (WS2) and cuprous oxide (Cu2O) are used as an electron transport layer (ETL) and hole transport layer (HTL), respectively. The PV performance parameters such as power conversion efficiency, open-circuit voltage (V oc), short-circuit current (J sc), and fill factor (FF) are assessed through adjustments in material properties including thickness, acceptor concentration, bulk defect density of the absorber, defect state of absorber/ETL and HTL/absorber interfaces, operating temperature, work function of the rear electrode, and cell resistances. This analysis aims to validate their collective impact on the overall efficiency of the designed Ni/Cu2O/Sb2(S,Se)3/WS2/FTO/Al TFSC. The optimized physical parameters for the Sb2(S,Se)3 TFSC lead to impressive PV outputs with an efficiency of 30.18%, V oc of 1.02 V, J sc of 33.65 mA cm-2, and FF of 87.59%. Furthermore, an artificial neural network (ANN) machine learning (ML) algorithm predicts the optimal PCE by considering five semiconductor parameters: absorber layer thickness, bandgap, electron affinity, electron mobility, and hole mobility. This model, which has an approximate correlation coefficient (R 2) of 0.999, is able to predict the data with precision. This numerical analysis provides valuable data for the fabrication of an environmentally friendly, economical, and incredibly non-toxic efficient heterojunction TFSC.

English translation evaluation method based on BP neural network

In order to solve the problems of machine translation efficiency and translation quality, this paper proposes an English translation evaluation system based on BP neural network algorithm. This method provides users with a more intelligent machine translation service experience. With the help of BP neural network algorithm, taking English online translation as the research object, Google's translation quality is the best, with an error frequency of only 167, while Baidu translation and iFLYTEK translation in China have a high error rate of 266 and 301 respectively, which is much higher than Google translation. A model of machine translation evaluation based on neural network algorithm is proposed to better solve the disadvantages of traditional English machine translation. The results show that the machine translation system based on neural network algorithm can further optimize the problems existing in machine translation, such as insufficient use of information and large scale of model parameters, and further improve the performance of neural network machine translation.

Multimodal Machine Translation

In recent years, neural network machine translation, especially in the field of multimodality, has developed rapidly. It has been widely used in natural languages processing tasks such as event detection and sentiment classification. The existing multimodal neural network machine translation is mostly based on the autoencoder framework of the attention mechanism, which further integrates spatial-visual features. However, due to the ubiquitous lack of corpus and the semantic interaction between multimodalities, the quality of machine translation is difficult to guarantee. Therefore, this paper proposes a multi-modal machine translation model that integrates external linguistic knowledge. Specifically, on the encoder side, we adopt the pre-trained Bert model to be used as an additional encoder to integrate with the original text encoder and picture encoder. Under the cooperation of the three encoders, a better text representation and picture representation at the source end is generated. Besides, the decoder decodes and generates a translation based on the image and text representation of the source. To sum up, this paper studies the visual-text semantic interaction on the encoder side and the visual-text semantic interaction on the decoder side, and further improves the quality of translation by introducing external linguistic knowledge. We compared the performance of the multimodal neural network machine translation model with pre-trained Bert and other baseline models in English German translation tasks on the multi30k data sets. The results show that the model can significantly improve the quality of multimodal neural network machine translation, which also verifies the importance of integrating external knowledge and visual text semantic interaction.

Possibilities and limitations of convolutional neural network machine learning architectures in the characterisation of achiral orthogonal smectic liquid crystals.

Machine learning is becoming a valuable tool in the characterisation and property prediction of liquid crystals. It is thus worthwhile to be aware of the possibilities but also the limitations of current machine learning algorithms. In this study we investigated a phase sequence of isotropic - fluid smecticA - hexatic smectic B - soft crystal CrE - crystalline. This is a sequence of transitions between orthogonal phases, which are expected to be difficult to distinguish, because of only minute changes in order. As expected, strong first order transitions such as the liquid to liquid crystal transition and the crystallisation can be distinguished with high accuracy. It is shown that also the hexatic SmB to soft crystal CrE transition is clearly characterised, which represents the transition from short- to long-range order. Limitations of convolutional neural networks can be observed for the fluid to hexatic SmA to SmB transition, where both phases exhibit short-range ordering.

A Coupled Machine Learning and Lattice Boltzmann Method Approach for Immiscible Two-Phase Flows

An innovative coupling numerical algorithm is proposed in the current paper, the front-tracking method–lattice Boltzmann method–machine learning (FTM-LBM-ML) method, to precisely capture fluid flow phase interfaces at the mesoscale and accurately simulate dynamic processes. This method combines the distinctive abilities of the FTM to accurately capture phase interfaces and the advantages of the LBM for easy handling of mesoscopic multi-component flow fields. Taking a single vacuole rising as an example, the input and output sets of the machine learning model are constructed using the FTM’s flow field, such as the velocity and position data from phase interface markers. Such datasets are used to train the Bayesian-Regularized Back Propagation Neural Network (BRBPNN) machine learning model to establish the corresponding relationship between the phase interface velocity and the position. Finally, the trained BRBPNN neural network is utilized within the multi-relaxation LBM pseudo potential model flow field to predict the phase interface position, which is compared with the FTM simulation. It was observed that the BRBPNN-predicted interface within the LBM exhibits a high degree of consistency with the FTM-predicted interface position, showing that the BRBPNN model is feasible and satisfies the accuracy requirements of the FT-LB coupling model.

Prediction of Skin Diseases using Convolutional Neural Networks as an Effort to Prevent Their Spread in Islamic Boarding School Environments

Skin disease is a common health problem in Islamic boarding school environments. This disease can spread quickly among students due to close contact and sharing the same facilities. Preventing the spread of skin diseases is a top priority in efforts to maintain the health and welfare of students in Islamic boarding schools. In this research, we propose the use of machine learning techniques to predict skin diseases in Islamic boarding school students. The main goal of this research is to develop a predictive model that can help identify skin diseases quickly and accurately. It is hoped that this will enable the prevention of the spread of skin diseases in the Islamic boarding school environment. The method used in this research involves the following steps: skin disease image data collection, data processing and cleaning, feature extraction from patient data, and machine learning model training and evaluation. We will use a Convolutional Neural Network (CNN) machine learning algorithm to build a predictive model. The dataset used in this research consists of images of melanoma, acne and acne skin diseases. In addition, validation will be carried out using data that has never been seen before to test the performance of the predictive model. It is hoped that the results of this research can make a significant contribution in preventing the spread of skin diseases in the Islamic boarding school environment. With accurate predictive models, health workers in Islamic boarding schools can take appropriate preventive measures to control skin diseases effectively. Apart from that, this research can also be a basis for developing a health information system that supports preventive measures for skin diseases in Islamic boarding schools more widely.

Development of NaCl-MgCl2-CaCl2 Ternary Salt for High-Temperature Thermal Energy Storage Using Machine Learning.

NaCl-MgCl2-CaCl2 eutectic ternary chloride salts are potential heat transfer and storage materials for high-temperature thermal energy storage. In this study, first-principles molecular dynamics simulation results were used as a data set to develop an interatomic potential for ternary chloride salts using a neural network machine learning method. Deep potential molecular dynamics (DPMD) simulations were performed to predict the microstructure and thermophysical properties of the NaCl-MgCl2-CaCl2 ternary salt. This work reveals that DPMD simulations can accurately calculate the microstructure and thermophysical properties of ternary chloride salts. The association strength of chloride ions and cations follows the order of Mg2+ > Ca2+ > Na+, and the coordination number decreases gradually with increasing temperature, indicating a progressively looser and more disordered molten structure. Furthermore, thermophysical properties, such as density, specific heat capacity, thermal conductivity, and viscosity, are in good agreement with the experimental measurements. Machine learning molecular dynamics will provide a feasible multivariate molten salt exploration method for the design of next-generation solar power plants and thermal energy storage systems.

Artificial intelligence-based automated model for prediction of extraction using neural network machine learning: A scope and performance analysis

To compare the Artificial Intelligence based model & conventional technique for prediction of extraction in orthodontic treatment plan. A comparative study was conducted on total 700 patients, who were divided into training set and testing set based on simple random sampling by means of computer generated random numbers. The photographs of the 630 patients [training set] along with the treatment plan finalized for them based on Arch Perimeter & Carey’s Analysis, was fed in the AI model [convolutional neural network (ResNet-50)] in order to train it for the stipulated function of eventually predicting the treatment plan in the testing set [70 patients], based on the input of the right profile photographs. The accuracy of measurement of the parameters of these seventy test set patients by the machine learning model relative to the manual method was compared eventually. Using the Statistical Package for Social Sciences, the acquired data was statistically analyzed, and p <0.05 was deemed statistically significant. The normality of the data was examined using the Shapiro-Wilks test and the Kolmogorov-Smirnov test. Depending on the collected data and normality assessed, appropriate reliability was estimated. The analysis of 70 test patients showed that 65.12% of the total extraction cases and 62.96% of the total non-extraction cases (as predicted by the AI model) were in agreement with the results of the model analysis.It is suggested that the present AI model can further be developed in order to improve the accuracy of prediction.

Transformer Neural Networks, Information Biology, and Alzheimer’s Disease

AbstractBackgroundAlzheimer’s disease affects one in ten people older than 65 years, with the prevalence expected to triple by 2030, together with commensurate increases in cost. A comprehensive causal model of Alzheimer’s disease etiology is lacking despite more than three decades of study. Modern computational methods provide a new approach to multiscalar biosystems by modeling neuropathology in its native three‐dimensional character in topological biophysics and quantum computational models.MethodThis work describes how an emerging standard in machine learning, responsible for the AlphaFold (protein folding structure) and Gato generalist agent projects, transformer neural networks, may be applied to the study of Alzheimer’s disease per proven results in bot text chat, image recognition and creation, video generation, software programming, and autonomy in robotics and driving. The self‐attention and self‐learning mechanisms of transformer neural networks are discussed in application to multifaceted Alzheimer’s disease data sets involving neuroimaging scans, whole human genome data, and blood and CSF biomarker data.ResultA machine learning‐based topological biophysics model of Alzheimer’s disease neuropathology hypothesis‐testing (misfolded proteins, gene expression, diabetes type‐3, and peptide immune system stimulation) is described.ConclusionThere is an opportunity to apply state‐of‐the‐art transformer neural network machine learning techniques to the automated analysis of large accruing data stores of Alzheimer’s information (imaging, genomic, proteomic, and self‐reported data) towards the causal understanding of the neuropathology.

Transformer Neural Networks, Information Biology, and Alzheimer’s Disease

AbstractBackgroundAlzheimer’s disease affects one in ten people older than 65 years, with the prevalence expected to triple by 2030, together with commensurate increases in cost. A comprehensive causal model of Alzheimer’s disease etiology is lacking despite more than three decades of study. Modern computational methods provide a new approach to multiscalar biosystems by modeling neuropathology in its native three‐dimensional character in topological biophysics and quantum computational models.MethodThis work describes how an emerging standard in machine learning, responsible for the AlphaFold (protein folding structure) and Gato generalist agent projects, transformer neural networks, may be applied to the study of Alzheimer’s disease per proven results in bot text chat, image recognition and creation, video generation, software programming, and autonomy in robotics and driving. The self‐attention and self‐learning mechanisms of transformer neural networks are discussed in application to multifaceted Alzheimer’s disease data sets involving neuroimaging scans, whole human genome data, and blood and CSF biomarker data.ResultA machine learning‐based topological biophysics model of Alzheimer’s disease neuropathology hypothesis‐testing (misfolded proteins, gene expression, diabetes type‐3, and peptide immune system stimulation) is described.ConclusionThere is an opportunity to apply state‐of‐the‐art transformer neural network machine learning techniques to the automated analysis of large accruing data stores of Alzheimer’s information (imaging, genomic, proteomic, and self‐reported data) towards the causal understanding of the neuropathology.

Machine Learning Model with Computed Tomography Radiomics and Clinicobiochemical Characteristics Predict the Subtypes of Patients with Primary Aldosteronism

Adrenal venous sampling (AVS) is the primary method for differentiating between primary aldosterone (PA) subtypes. The aim of study is to develop prediction models for subtyping of patients with PA using computed tomography (CT) radiomics and clinicobiochemical characteristics associated with PA. This study retrospectively enrolled 158 patients with PA who underwent AVS between January 2014 and March 2021. Neural network machine learning models were developed using a two-stage analysis of triple-phase abdominal CT and clinicobiochemical characteristics. In the first stage, the models were constructed to classify unilateral or bilateral PA; in the second stage, they were designed to determine the predominant side in patients with unilateral PA. The final proposed model combined the best-performing models from both stages. The model's performance was evaluated using repeated stratified five-fold cross-validation. We employed paired t-tests to compare its performance with the conventional imaging evaluations made by radiologists, which categorize patients as either having bilateral PA or unilateral PA on one side. In the first stage, the integrated model that combines CT radiomic and clinicobiochemical characteristics exhibited the highest performance, surpassing both the radiomic-alone and clinicobiochemical-alone models. It achieved an accuracy and F1 score of 80.6%±3.0% and 74.8%±5.2% (area under the receiver operating curve [AUC] = 0.778±0.050). In the second stage, the accuracy and F1 score of the radiomic-based model were 88%±4.9% and 81.9%±6.2% (AUC=0.831±0.087). The proposed model achieved an accuracy and F1 score of 77.5%±3.9% and 70.5%±7.1% (AUC=0.771±0.046) in subtype diagnosis and lateralization, surpassing the accuracy and F1 score achieved by radiologists' evaluation (p<.05). The proposed machine learning model can predict the subtypes and lateralization of PA. It yields superior results compared toconventional imaging evaluation and has potential to supplement the diagnostic process in PA.

Predicting Groundwater Level using Temporal Attention Enhanced Graph Neural Network

A revolutionary approach for groundwater management is essential. To predicting the water levels using for integrating IoT sensors, cloud computing, and advanced data analysis methods. IoT sensors are employed for real-time measurement of groundwater levels, creating a robust dataset. The paper focuses on predicting future groundwater levels, a crucial aspect of sustainable resource management. To enhance predictive accuracy, a preprocessing algorithm, such as Min-Max normalization, is introduced to clean and normalize the collected data, ensuring its reliability. Additionally, a feature extraction algorithm, such as Principal Component Analysis (PCA), is implemented to identify relevant patterns and trends within the dataset, enhancing the efficiency of subsequent analysis. A novel classification algorithm, Spatial-Temporal Graph Convolutional Network, is introduced, enabling the identification of potential groundwater recharge areas. This classification algorithm leverages historical data and extracted features to categorize regions based on their suitability for groundwater renewal. Finally, this research uses a Temporal attention-enhanced graph Neural Network machine learning algorithm to predict groundwater levels in the next few years. This algorithm utilizes the preprocessed data and extracted features, identifying intricate patterns and trends in historical data to generate precise predictions for groundwater levels in the upcoming years.

A network-based approach to improving robustness of a high-speed train by structure adjustment

To enhance the ability of a high-speed train (HST) to provide services under various adverse conditions, this paper studies robustness improvement of the HST by adjusting its structure. We model the HST as an interdependent machine-electricity-communication network (IMECN) composed of a machine network (MN), an electricity network (EN) and a communication network (CN), and propose a robustness metric of the IMECN subject to node failures considering failure propagation. Then, a robustness optimization model is constructed for structure adjustment. A Tabu search algorithm with directed and undirected edge exchange operators is designed to solve this problem. A case study on a practical HST is used to verify the feasibility and effectiveness of the proposed method. The results show that robustness of the IMECN is significantly improved by structure adjustment. Furthermore, most nodes have little effect on robustness, and impact of failures on any node is minimized after optimization. In addition, in terms of topology, nodes with low and high degrees in the MN and CN reconnect with those with a similar degree, their clustering coefficients become larger, and closeness of all nodes in the subnetworks increases. Finally, adjusting the structures leads to a slight difference of loads in the EN and CN.

Home Economics Teachers' Concern and Perception about Home Economics Education Using the Latest Technology in the Era of the 4th Industrial Revolution

The purpose of this study was to identify home economics (HE) teachers' concerns about and perceptions of HE education using the latest technologies in the era of the 4th Industrial Revolution and to reveal whether they differ according to teachers' general background variables. The questionnaire survey method to measure HE teachers’ concerns and perceptions of HE education using the latest technologies in the era of the 4th Industrial Revolution was conducted online using the Google Questionnaire from which 150 responses were received. The main results were as follows. Firstly, HE teachers scored an average of 3.46 out of 5 for the latest technology. Among these interests in the latest technology, interest in “augmented reality and virtual reality technologies” scored the highest at an average of 3.80, while interest in “neural network machine learning” (2.78) was low. HE teacher's concerns about HE education using the latest technologies in the era of the 4th Industrial Revolution were high, with an average score of 4.40. Among these concerns for the latest technology, “concern about the results of HE education using the latest technology” scored the highest at 4.53. HE teachers’ anxiety about the latest teaching technology in the era of the 4th Industrial Revolution was moderate, averaging 3.05. The highest form of anxiety was “anxiety about the impact on the job” (4.03) and the lowest was fear of “the disappearance of the teacher's job” (2.50). HE teachers' innovation resistance to the latest teaching technology was low at 2.18. Expectations of the latest technology in HE classes in the era of the 4th Industrial Revolution averaged 3.85, slightly higher than the middle of 3.

Analysis method of apparent quality of fair-faced concrete based on convolutional neural network machine learning

Fair-faced concrete has attracted much attention because of its noble texture, but the evaluation of its appearance quality often depends on personal subjective feelings (especially color difference) or complex image processing means, which hinders its development. Therefore, the accurate and objective evaluation of appearance quality is the key to fair-faced concrete engineering. This paper proposed a machine learning method to analyze the appearance quality of fair-faced concrete. Firstly, images of fair-faced concrete were collected by UAV (Unmanned Aerial Vehicle) or scanner for pretreatment. Then, based on the convolutional neural network algorithm, a visual quality database of fair-faced concrete including 2463 groups of information was established, and it was trained to recognize color difference and bubbles on the surface of fair-faced concrete pictures. The performance of the model is evaluated and optimized. The influence of environmental factors such as light intensity and surface moisture on the model is analyzed. Finally, the machine learning method is used to analyze the appearance quality of fair-faced concrete in engineering site. Compared with the traditional analysis methods, the accuracy of stomatal area rate and maximum stomatal diameter recognition was increased from 79.34 % to 93.83 %, and the accuracy of color difference recognition was increased by more than 10 %. At the same time, the identification method in this study still has strong robustness under light and surface moisture variation. The research results can provide useful thinking for evaluating fair-faced concrete engineering.