Variational Autoencoder Model Research Articles

Just-in-time framework for robust soft sensing based on robust variational autoencoder

Modeling with high-dimensional data subject to abnormal observations have always been a practical interest. In this paper, under the just-in-time learning (JITL) framework, a robust soft sensor modeling approach is developed based on robust Variational Autoencoder (VAE). Unlike the vanilla VAE that extracts features from the given dataset under the Gaussian prior assumption, robust VAE employs Student’s t-distribution as prior distribution to handle abnormal data. Under assumption of the Student’s t-prior, the proposed robust VAE model is capable of describing collected data contaminated with outliers. Once the robust VAE model is trained, each robust feature variable in the latent space can be determined. Subsequently, similarity measure is calculated using robust Kullback-Leibler divergence between two Student’s t-distributions, that is, the distribution of a new data sample and that of each historical data sample. After completing similarity measurement for a query sample, the weights for input-output historical data can be determined. Based on these weighted historical data samples, a robust probabilistic principal component regression (PPCR) is utilized to perform local modeling for prediction. Numerical simulations, including the Tennessee Eastman and Penicillin fermentation benchmark processes, are utilized to validate the proposed JITL-based robust soft sensor modeling method.

Machine learning assisted crystallographic reconstruction from atom probe tomographic images

Atom probe tomography (APT) is a powerful technique for three-dimensional (3D) atomic-scale imaging, enabling the accurate analysis on the compositional distribution at the nanoscale. How to accurately reconstruct crystallographic information from APT data, however, is still a great challenge due to the intrinsic nature of the APT technique. In this paper, we propose a novel approach that consists of the modified forward simulation process and the backward machine learning process to recover the tested crystal from APT data. The high-throughput forward simulations on Al single crystals of different orientations generate 10 000 original 3D images and data augmentation is implemented on the original images, resulting in 100 000 3D images. The big data allows the development of deep learning models and three deep learning algorithms of Convolutional Neural Network (CNN), Vision Transformer (ViT), and Variational Autoencoder (VAE) are used in the backward process. After training, the ViT model performs superior than the CNN and VAE models, which can recover the crystalline orientation outstandingly, as evaluated by the coefficient of determinationR2and the Mean Percent Error (MPE), viz.,R2= 0.93 and MPE = 0.43%,R2= 0.97 and MPE = 0.35%, andR2= 0.93 and MPE = 0.77% for the rotation anglesϕ,ψandθ, respectively, on the test dataset. The present work clearly demonstrates the capability of deep learning models in the recovery of the tested crystals from APT data, thereby paving the way for the further development of large artificial intelligent models of APT.

Airfoil Shape Generation and Feature Extraction Using the Conditional VAE-WGAN-gp

A machine learning method was applied to solve an inverse airfoil design problem. A conditional VAE-WGAN-gp model, which couples the conditional variational autoencoder (VAE) and Wasserstein generative adversarial network with gradient penalty (WGAN-gp), is proposed for an airfoil generation method, and then, it is compared with the WGAN-gp and VAE models. The VAEGAN model couples the VAE and GAN models, which enables feature extraction in the GAN models. In airfoil generation tasks, to generate airfoil shapes that satisfy lift coefficient requirements, it is known that VAE outperforms WGAN-gp with respect to the accuracy of the reproduction of the lift coefficient, whereas GAN outperforms VAE with respect to the smoothness and variations of generated shapes. In this study, VAE-WGAN-gp demonstrated a good performance in all three aspects. Latent distribution was also studied to compare the feature extraction ability of the proposed method.

Unsupervised electrocardiogram feature extraction using deep learning empowers discovery of genetic and phenotypic determinants of cardiac electrophysiology

Abstract Background Conventional analysis of the electrocardiogram (ECG) is based on the extraction of human-defined, visually recognisable features. However, these are not optimised to capture all the information content in the ECG. The variational autoencoder (VAE), a form of unsupervised machine learning, can address this shortcoming by computationally extracting comprehensive and interpretable new ECG features, called latent factors. These latent factors provide a low-dimensional representation of the ECG that maximises capture of data content. This approach could expand our understanding of electrophysiology by identifying novel genetic and phenotypic traits associations with the ECG. Purpose This study aims to uncover genetic determinants of VAE latent factors, comparing them to determinants of conventional, human-derived ECG parameters to gain novel insights into cardiac electrophysiology and related diseases. Methods Over one million median beat ECGs derived from a United States (US) based secondary care centre were used to train a machine learning VAE model, with external validation in the UK Biobank (UKB). We performed common and rare variant association studies for VAE latent factors and conventional ECG traits on quality-controlled UKB data. Associated genetic variants were compared to loci for conventional ECG parameters available in the UKB and literature. Novel GWAS associations were validated in a withheld subset of the UKB cohort. Additionally, we compared the associations of the VAE latent factors and conventional ECG traits with phenotypic traits, disease codes and echocardiographic traits. Results Utilising median beat ECGs, the VAE model identified 20 features, with partial correlations to traditional ECG traits. A genome-wide association study (GWAS) identified 105 associated genomic regions, compared with 62 regions identified with conventional ECG traits on the same dataset. Most discovered loci have been previously associated with ECG features in larger GWAS, providing a positive control validation. We also discovered and validated five genomic regions not previously linked to the ECG (mapped to TRIOBP, EFEMP1, NEDD9, ATP10A and P2RX1). We found rare variant associations for seven genes (NEK6, IL17RA, NME7, MYBPC3, CCT8, ADAMTS6 and SCN5A). The latent factors uncovered associations with 158 phenotypic traits (2093 phenotypes tested) and 147 disease codes (2074 disease phecodes tested) which were not identified by ECG traits. Out of 38 echocardiographic traits, 35 showed higher correlations with the latent factors, while six exhibited significant associations exclusively with the latent factors. Conclusion Our study demonstrates that VAE-derived latent factors outperform conventional ECG parameters in uncovering novel genetic and phenotypic associations. Our findings underscore the opportunity for optimised data-driven feature extraction to enhance our understanding of cardiac electrophysiology.

Convolutional variational autoencoder and multi-scale attention convolutional neural network based diagnostics on filament current sensors for mass spectrometers

Fault detection of the filament current sensor of the spaceborne mass spectrometer is of great significance to the safe operation of the mass spectrometer. Neglecting sensor failures may affect the normal operation of the mass spectrometer, which in turn affects the health of the astronauts and the space missions. This paper proposes an enhanced intelligent diagnosis method based on filament current sensor for mass spectrometer via improved deep learning network, aiming to improve the accuracy and reliability of the filament current sensor when the fault samples are insufficient. The improved convolutional variational autoencoder (CVAE) model not only enhances the data for four typical sensor fault signals, namely bias, drift, missing and random, but also solves the problem of insufficient samples when spike, Precision degradation and stuck fault occur in sensors. Short-time Fourier transform (STFT) is utilized to transform one-dimensional data into two-dimensional spectrograms, which gives more fault characteristics to the data set. The improved multi-scale attention mechanism convolutional neural network (MSAM-CNN) model is constructed to perform fault diagnosis on the generated spectrogram, which solves the problem of low accuracy of fault identification in traditional convolutional neural network (CNN) models. The results of ablation and comparison experiments show that the samples generated by CVAE have smaller Mean Absolute Error (MAE), Mean Square Error (MSE), and Root Mean Square Error (RMSE), the enhanced samples improve the classification and clustering of MSAM-CNN, and compared to other diagnostic models, MSAM-CNN achieves the highest accuracy, precision, recall, and F1-score of 99.2%, 99.3%, 98.8%, and 0.991.

Deep Learning with Crested Porcupine Optimizer for Detection and Classification of Paddy Leaf Diseases for Sustainable Agriculture

India has a vast number of inhabitants and the main food source distribution is from agriculture. Agricultural lands are being demolished generally owing to plant and crop illnesses. The detection of plant diseases by using image processing models can aid agriculturalists in defending the farming area from damaging or affecting it. Paddy is the main harvest worldwide. Early recognition of the paddy diseases at dissimilar phases of development is very vital in paddy production. However, the present manual technique in identifying and classifying paddy diseases needs a very educated farmer and is time-consuming. Deep learning (DL) is an effectual research area in the classification of agriculture patterns where it can efficiently solve the problems of diseases identification. Therefore, the articles focus on the design and expansion of Deep Learning based Crested Porcupine Optimizer for the Detection and Classification of Paddy Leaf Diseases (DLCPO-DCPLD) method for Sustainable Agriculture. The main aim of the DLCPO-DCPLD method use DL method for the recognition and identification of rice plant leaf diseases. To accomplish this, the DLCPO-DCPLD technique performs the image pre-processing using Median Filtering (MF) to recover the excellence of the input frames. Next, the ConvNeXt-L method is applied for extraction of feature vectors from the pre-processed images. Also, the Conditional Variational Autoencoder (CVAE) model is utilized for the automated classification of Paddy Leaf diseases. Eventually, the hyperparameter tuning of the CVAE technique is accomplished by implementing the Crested Porcupine Optimizer (CPO) technique. To safeguard the enhanced predictive results of the DLCPO-DCPLD method, a sequence of experimentations is implemented on the benchmark dataset. The experimental validation of the DLCPO-DCPLD method portrayed a superior accuracy value of 99.12% over existing approaches.

Variational Autoencoder with Gaussian Random Field prior: Application to unsupervised animal detection in aerial images

In real world datasets of aerial images, the objects of interest are often missing, hard to annotate and of varying aspects. The framework of unsupervised Anomaly Detection (AD) is highly relevant in this context, and Variational Autoencoders (VAEs), a family of popular probabilistic models, are often used. We develop on the literature of VAEs for AD in order to take advantage of the particular textures that appear in natural aerial images. More precisely we propose a new VAE model with a Gaussian Random Field (GRF) prior (VAE-GRF), which generalizes the classical VAE model, and we provide the necessary procedures and hypotheses required for the model to be tractable. We show that, under some assumptions, the VAE-GRF largely outperforms the traditional VAE and some other probabilistic models developed for AD. Our results suggest that the VAE-GRF could be used as a relevant VAE baseline in place of the traditional VAE with very limited additional computational cost. We provide competitive results on the MVTec reference dataset for visual inspection, and two other datasets dedicated to the task of unsupervised animal detection in aerial images.

Uncovering neural substrates across Alzheimer's disease stages using contrastive variational autoencoder.

Alzheimer's disease is the most common major neurocognitive disorder. Although currently, no cure exists, understanding the neurobiological substrate underlying Alzheimer's disease progression will facilitate early diagnosis and treatment, slow disease progression, and improve prognosis. In this study, we aimed to understand the morphological changes underlying Alzheimer's disease progression using structural magnetic resonance imaging data from cognitively normal individuals, individuals with mild cognitive impairment, and Alzheimer's disease via a contrastive variational autoencoder model. We used contrastive variational autoencoder to generate synthetic data to boost the downstream classification performance. Due to the ability to parse out the nonclinical factors such as age and gender, contrastive variational autoencoder facilitated a purer comparison between different Alzheimer's disease stages to identify the pathological changes specific to Alzheimer's disease progression. We showed that brain morphological changes across Alzheimer's disease stages were significantly associated with individuals' neurofilament light chain concentration, a potential biomarker for Alzheimer's disease, highlighting the biological plausibility of our results.

Deep-Learning-Based Disease Classification in Patients Undergoing Cine Cardiac MRI.

Automated approaches may allow for fast, reproducible clinical assessment of cardiovascular diseases from MRI. To develop an MRI-based deep learning (DL) disease classification algorithm to distinguish among normal subjects (NORM), patients with dilated cardiomyopathy (DCM), hypertrophic cardiomyopathy (HCM), and ischemic heart disease (IHD). Retrospective. A total of 1337 subjects (55% female), comprising normal subjects (N = 568), and patients with DCM (N = 151), HCM (N = 177), and IHD (N = 441). Balanced steady-state free precession cine sequence at 1.5/3.0 T. Bi-ventricular morphological and functional features and global and segmental left ventricular strain features were automatically extracted from short- and long-axis cine images. Variational autoencoder models were trained on the extracted features and compared against consensus disease label provided by two expert readers (13 and 14 years of experience). Adding unlabeled, normal data to the training was explored to increase specificity of NORM class. Tenfold cross-validation for model development; mean, standard deviation (SD) for measurements; classification metrics: area under the curve (AUC), confusion matrix, accuracy, specificity, precision, recall; 95% confidence intervals; Mann-Whitney U test for significance. AUCs of 0.952 for NORM, 0.881 for DCM, 0.908 for HCM, and 0.856 for IHD and overall accuracy of 0.778 were obtained, with specificity of 0.908 for the NORM class using both SAX and LAX features. Longitudinal strain features slightly improved classification metrics by 0.001 to 0.03 points, except for HCM-AUC. Differences in accuracy, metrics for NORM class and HCM-AUC were statistically significant. Cotraining using unlabeled data increased the specificity for the NORM class to 0.961. Cardiac function features automatically extracted from cine MRI have potential to be used for disease classification, especially for normal-abnormal classification. Feature analyses showed that strain features were important for disease labeling. Cotraining using unlabeled data may help to increase specificity for normal-abnormal classification. 3 TECHNICAL EFFICACY: Stage 1.

The Defect Identification System of Electromechanical Equipment on the Edge Side of the Power Grid under Edge Computing

With the development of the industrial Internet of Things, modern industrial systems have developed towards intelligence. Electromechanical Equipment (EE) is essential, and its defect identification is fundamental. Firstly, this research introduces the basic content of Gated Recurrent Unit (GRU), Variational Auto-Encoder (VAE) in Deep Learning (DL), and Edge Computing (EC) to explore the construction of a defect identification system for EE on the edge side of the power grid. Secondly, combined with the advantages of GRU and VAE, a GRU-VAE defect recognition model is proposed. Then, the EC architecture is introduced, and the EE defect identification system based on the GRU-VAE algorithm is constructed. The EC intelligent EE defect identification service system is designed with this as the core. Finally, simulation experiments are carried out using different data sets to verify the performance of the GRU-VAE model. The results show that the GRU-VAE model has higher precision and recall than the separate GRU model and VAE model, and the corresponding F1 value is also higher. The F1 value can reach 0.997 on aperiodic data and 0.966 on periodic data. In addition, the optimal thresholds of different datasets are analyzed, and the relationship between the length of the time window and the model’s performance is studied. When the time window length is 15, the model performance is optimal. This research on the defect identification system of EE on the edge side of the power grid based on EC and DL can provide a new path and inject new vitality into the defect detection of EE.

Classification of the L-, H-mode, and plasma-free state: Convolutional neural networks and variational autoencoders on the edge reflectometer for KSTAR.

Classifying and monitoring the L-, H-mode, and plasma-free state are essential for the stable operational control of tokamaks. Edge reflectometry measures plasma density profiles, but the large volume of data and complexity in reconstruction pose significant challenges. There is a need for efficient methods to analyze complex reflectometer data in real-time, which can be addressed using advanced computational techniques. Here, we show that machine learning (ML) techniques can classify discharge states using raw signal data from an edge reflectometer installed on the Korea Superconducting Tokamak Advanced Research. The deep convolutional neural network models achieved classification accuracy of up to 99% when using 2D spectrogram inputs, demonstrating a significant improvement over 1D raw signal inputs. Additionally, the variational autoencoder model effectively clustered the discharge states in the latent space without any label information, further validating the model's capability to classify discharge states. These results suggest that the ML model can effectively handle the complexity of reflectometer data and accurately classify plasma discharge states. This approach not only facilitates real-time diagnosis but also reduces the need for manual data processing.

Inhibitor_Mol_VAE: a variational autoencoder approach for generating corrosion inhibitor molecules

Deep learning-based generative modeling demonstrates proven advantages as an effective approach in molecular discovery. This study introduces a generative-network based method called Inhibitor_Mol_VAE, which uses a variational autoencoder model to generate corrosion inhibitor molecules with targeted inhibition efficiency. We first evaluate the model’s ability to reconstruct molecules. Then, we assess the model’s ability to generate new inhibitor molecules using physiochemical properties (including MolWt, LogP, Vdw_volume, and Electronegativity). New molecules with high inhibition efficiencies at low concentrations, such as [ethoxy(methoxy)phosphoryl]-phenylmethanol and (alpha-methylamino-benzyl)-phosphonsaeure-monoaethylester are successfully discovered.

Efficient property-oriented design of composite layups via controllable latent features using generative VAE

Fiber-reinforced composites provide substantial tailoring potential, while the extensive parameters and complex coupling mechanisms pose formidable challenges to layup designs. This paper presents an efficient inverse design framework for composite layups utilizing a variational autoencoder (VAE), which is applicable to non-conventional laminates. By leveraging the VAE's exceptional feature extraction and generative capabilities, the decoder rapidly produces layups with desired properties through controllable feature vectors. Based on the stacking characteristics of layups, multi-scale one-dimensional convolutions precisely extract sequence features relevant to mechanical properties and specific manufacturing constraints. A customized loss function is formulated to constrain the latent features, while addressing the non-uniqueness problem for layups with certain mechanical properties. The developed property-oriented VAE can generate 100,000 layups in seconds, achieving an average success rate of 66.9% under comprehensive in-plane and bending stiffness design, and remains effective for 100-ply thick laminate. For comparison, the VAE model outperforms the genetic algorithm and the logic-based method in reinforced panel designs, reducing the retrieval error by 46.4% and 38.1%, respectively. The proposed approach demonstrates flexible and efficient design advantages using generative machine learning models, and is easily extendable to other inverse design scenarios.

ScVGATAE: A Variational Graph Attentional Autoencoder Model for Clustering Single-Cell RNA-seq Data.

Single-cell RNA sequencing (scRNA-seq) is now a successful technology for identifying cell heterogeneity, revealing new cell subpopulations, and predicting developmental trajectories. A crucial component in scRNA-seq is the precise identification of cell subsets. Although many unsupervised clustering methods have been developed for clustering cell subpopulations, the performance of these methods is prone to be affected by dropout, high dimensionality, and technical noise. Additionally, most existing methods are time-consuming and fail to fully consider the potential correlations between cells. In this paper, we propose a novel unsupervised clustering method called scVGATAE (Single-cell Variational Graph Attention Autoencoder) for scRNA-seq data. This method constructs a reliable cell graph through network denoising, utilizes a novel variational graph autoencoder model integrated with graph attention networks to aggregate neighbor information and learn the distribution of the low-dimensional representations of cells, and adaptively determines the model training iterations for various datasets. Finally, the obtained low-dimensional representations of cells are clustered using kmeans. Experiments on nine public datasets show that scVGATAE outperforms classical and state-of-the-art clustering methods.

IEV2Mol: Molecular Generative Model Considering Protein-Ligand Interaction Energy Vectors.

Generating drug candidates with desired protein-ligand interactions is a significant challenge in structure-based drug design. In this study, a new generative model, IEV2Mol, is proposed that incorporates interaction energy vectors (IEVs) between proteins and ligands obtained from docking simulations, which quantitatively capture the strength of each interaction type, such as hydrogen bonds, electrostatic interactions, and van der Waals forces. By integrating this IEV into an end-to-end variational autoencoder (VAE) framework that learns the chemical space from SMILES and minimizes the reconstruction error of the SMILES, the model can more accurately generate compounds with the desired interactions. To evaluate the effectiveness of IEV2Mol, we performed benchmark comparisons with randomly selected compounds, unconstrained VAE models (JT-VAE), and compounds generated by RNN models based on interaction fingerprints (IFP-RNN). The results show that the compounds generated by IEV2Mol retain a significantly greater percentage of the binding mode of the query structure than those of the other methods. Furthermore, IEV2Mol was able to generate compounds with interactions similar to those of the input compounds, regardless of structural similarity. The source code and trained models for IEV2Mol, JT-VAE, and IFP-RNN designed for generating compounds active against the DRD2, AA2AR, and AKT1, as well as the data sets (DM-QP-1M, active compounds to each protein, and ChEMBL33) utilized in this study, are released under the MIT License and available at https://github.com/sekijima-lab/IEV2Mol.

Accelerated discovery of eutectic compositionally complex alloys by generative machine learning

Eutectic alloys have garnered significant attention due to their promising mechanical and physical properties, as well as their technological relevance. However, the discovery of eutectic compositionally complex alloys (ECCAs) (e.g. high entropy eutectic alloys) remains a formidable challenge in the vast and intricate compositional space, primarily due to the absence of readily available phase diagrams. To address this issue, we have developed an explainable machine learning (ML) framework that integrates conditional variational autoencoder (CVAE) and artificial neutral network (ANN) models, enabling direct generation of ECCAs. To overcome the prevalent problem of data imbalance encountered in data-driven ECCA design, we have incorporated thermodynamics-derived data descriptors and employed K-means clustering methods for effective data pre-processing. Leveraging our ML framework, we have successfully discovered dual- or even tri-phased ECCAs, spanning from quaternary to senary alloy systems, which have not been previously reported in the literature. These findings hold great promise and indicate that our ML framework can play a pivotal role in accelerating the discovery of technologically significant ECCAs.

SM-GMVAE: an intelligent model for defect quantification evaluation based on few ultrasonic signals

The conventional defect quantification evaluation approaches based on machine learning requires massive amounts of labelled defect signals, which is expensive and time-consuming works. This paper proposed a novel Similarity Metric Gaussian Mixture Variational Auto-Encoder (SM-GMVAE) model, which enables quantify defect with few labelled defect signals. The SM-GMVAE model is designed based on few-shot learning, which includes two modules: feature extraction (FE) module and similarity metric (SM) module. The FE module is designed to extract the feature of defect signal via the Variational Auto-Encoder (VAE). The SM module is used to measure the similarity of two defect signals based on the Gaussian Mixture Model (GMM). Moreover, sparse filtering techniques are used to enhance the sparsity of the features in the SM module. To validate proposed model, some specimens with four various depth defects are designed and fabricated for ultrasonic non-destructive testing experiments. A dataset with defects of different depths is established to compare proposed model with other methods. Our method obtains state-of-the-art experimental results with few labelled defect signals. Different from many published papers, our model is trained with few labelled data, which is more close to engineering practical application than other evaluation model trained using large numbers of labelled data. In other words, the developed approach can realize more complex defect evaluation tasks (such as: size, location, shapes, etc) at very low data labelling cost.

Prairie Dog Optimization Algorithm with deep learning assisted based Aerial Image Classification on UAV imagery

This study presents a Prairie Dog Optimization Algorithm with a Deep learning-assisted Aerial Image Classification Approach (PDODL-AICA) on UAV images. The PDODL-AICA technique exploits the optimal DL model for classifying aerial images into numerous classes. In the presented PDODL-AICA technique, the feature extraction procedure is executed using the EfficientNetB7 model. Besides, the hyperparameter tuning of the EfficientNetB7 technique uses the PDO model. The PDODL-AICA technique uses a convolutional variational autoencoder (CVAE) model to detect and classify aerial images. The performance study of the PDODL-AICA model is implemented on a benchmark UAV image dataset. The experimental values inferred the authority of the PDODL-AICA approach over recent models in terms of dissimilar measures.

Variational Autoencoders for Network Lifetime Enhancement in Wireless Sensors.

Wireless sensor networks (WSNs) are structured for monitoring an area with distributed sensors and built-in batteries. However, most of their battery energy is consumed during the data transmission process. In recent years, several methodologies, like routing optimization, topology control, and sleep scheduling algorithms, have been introduced to improve the energy efficiency of WSNs. This study introduces a novel method based on a deep learning approach that utilizes variational autoencoders (VAEs) to improve the energy efficiency of WSNs by compressing transmission data. The VAE approach is customized in this work for compressing WSN data by retaining its important features. This is achieved by analyzing the statistical structure of the sensor data rather than providing a fixed-size latent representation. The performance of the proposed model is verified using a MATLAB simulation platform, integrating a pre-trained variational autoencoder model with openly available wireless sensor data. The performance of the proposed model is found to be satisfactory in comparison to traditional methods, like the compressed sensing technique, lightweight temporal compression, and the autoencoder, in terms of having an average compression rate of 1.5572. The WSN simulation also indicates that the VAE-incorporated architecture attains a maximum network lifetime of 1491 s and suggests that VAE could be used for compression-based transmission using WSNs, as its reconstruction rate is 0.9902, which is better than results from all the other techniques.

Dynamic Soft Sensor Model for Endpoint Carbon Content and Temperature in BOF Steelmaking Based on Adaptive Feature Matching Variational Autoencoder

The key to endpoint control in basic oxygen furnace (BOF) steelmaking lies in accurately predicting the endpoint carbon content and temperature. However, BOF steelmaking data are complex and change distribution due to variations in raw material batches, process adjustments, and equipment conditions, leading to concept drift and affecting model performance. In order to resolve these problems, this paper proposes a dynamic soft sensor model based on an adaptive feature matching variational autoencoder (VAE-AFM). Firstly, this paper innovatively proposes an adaptive feature matching (AFM) method. This method utilizes the maximum mean discrepancy to calculate the values of the marginal and conditional distributions. Based on the discrepancy between these two values, a dynamic adjustment algorithm is designed to adaptively assign different weights to the two distributions. This approach dynamically and quantitatively evaluates and adjusts the relative importance of different distributions in the domain adaptation process, thereby enhancing the effectiveness of cross-domain data alignment. Secondly, a variational autoencoder (VAE) is employed to process the data, as the VAE model can capture the complex data structures and latent features in the steelmaking process. Finally, the features extracted by the VAE are processed with the adaptive feature matching method, thereby constructing the VAE-AFM dynamic soft sensor model. Experimental studies on actual BOF steelmaking data validate the efficacy of the offered approach, offering a reliable solution to the challenges of high complexity and concept drift in BOF steelmaking data.