autoencoder-based anomaly detectors generally utilize reconstruction error as an anomaly score. However, when both normal and abnormal data are included in train dataset, sufficient training time can lead to the learn...
详细信息
autoencoder-based anomaly detectors generally utilize reconstruction error as an anomaly score. However, when both normal and abnormal data are included in train dataset, sufficient training time can lead to the learning of both types of data. Therefore, it becomes difficult to distinguish between the two classes based on reconstruction errors, which degrades the detection performance. To mitigate this, it is necessary to regularize the number of training epochs (NTE). However, this is difficult without labels. Therefore, some studies introduced additional hyperparameters (HPs), arbitrarily selected, to determine the NTE. However, this may degrade the performance and increase its variability. To address this, cumulative error scoring (CES) has been proposed. This method does not determine a single optimal NTE. Instead, to enhance the robustness, it accumulates reconstruction errors from the models trained in the early stages of learning, similar to an ensemble method. Therefore, it showed promising results in the performance robustness. However, it also introduced an additional HP to exclude meaningless reconstruction errors. To improve the limitations of CES, we proposed adaptively weighted-cumulative error scoring (AW-CES). AW-CES utilizes the weighted average of reconstruction errors as an anomaly score. The weights are estimated based on the theoretical characteristic of inlier priority (IP) in deep learning models. Moreover, AW-CES does not require additional HP and the weights are automatically determined during the training without labels. The performance was evaluated using various experiments on real datasets. The results showed an improved average performance. In particular, AW-CES demonstrated a significant improvement in the performance robustness.
The quality of edible oils is closely related to their chemical compositions. Antioxidants have widespread application in edible oil production. In this study, a pioneering detection approach involving the use of a on...
详细信息
The quality of edible oils is closely related to their chemical compositions. Antioxidants have widespread application in edible oil production. In this study, a pioneering detection approach involving the use of a onedimensional convolutional autoencoder (1D-CAE) was introduced to compress spectral data for assessing antioxidant levels in edible oils. Fourier-transform near-infrared (FT-NIR) characterisation of edible oil samples with varying antioxidant concentrations was also conducted. An 1D-CAE model was developed to compress different pre-processed spectra into a condensed representation. These compressed features were then integrated with a support vector machine and partial least squares regression models to establish correlations for each target. The study examined the influence of pre-processing steps and feature engineering methods on near-infrared spectral analysis through independent or combined model analysis. The findings revealed that features derived from the 1D-CAE model demonstrated remarkable repeatability and can be utilised to construct robust detection models. The experimental results showed that the optimal detection model derived based on the 1D-CAE compression features has an average R2, 2 , RPD and RMSE of 0.9953, 15.1664 and 1.2035, respectively, on the prediction set. FT-NIR spectroscopy can be used to accurately detect butylated hydroxytoluene in edible oils. Therefore, autoencoders are an effective tool in spectroscopic analysis, offering promising avenues for future research and application.
Despite recent advances in single-image-based 3D human pose and shape estimation, partial occlusion remains a major challenge for many methods, leading to significant prediction errors. Some existing methods fail to p...
详细信息
ISBN:
(纸本)9789819984312;9789819984329
Despite recent advances in single-image-based 3D human pose and shape estimation, partial occlusion remains a major challenge for many methods, leading to significant prediction errors. Some existing methods fail to provide satisfactory performance for 3D human body reconstruction in occluded outdoor environments. To address these issues, we propose an autoencoder for feature extraction that integrates image masking methods to improve training stability. Our approach utilizes an attention mechanism to effectively capture the features of partially visible body parts, addressing partial occlusion. We further employ a partial attention mechanism to obtain the final features and use a regressor to estimate human model parameters. Experimental results on outdoor 3D poses in benchmark datasets demonstrate that our method outperforms state-of-the-art image-based methods in terms of robustness and efficiency. Qualitative evaluation shows that our method achieves more accurate and robust reconstruction results than existing methods, not only in occluded scenarios but also on standard benchmarks. Our approach exhibits excellent model robustness and training stability.
The anomaly detection of widely distributed sensors in small modular reactors (SMRs) is critical to the safe operation of systems. Existing methods of sensor anomaly detection for transient conditions are less studied...
详细信息
The anomaly detection of widely distributed sensors in small modular reactors (SMRs) is critical to the safe operation of systems. Existing methods of sensor anomaly detection for transient conditions are less studied and have some shortcomings, necessitating further research in this context. Therefore, this paper presents an unsupervised deep-learning framework for a gated recurrent autoencoder with the Luong attentional mechanism and the residual connection, which is a multiple-input and multiple-output model. A single model can monitor SMR systems equipped with many sensors, and the training of the model does not require an anomaly label. The general threshold-setting method of the autoencoder for anomaly detection is also improved. The Bayesian estimation algorithm sets a dynamic threshold value in the residual evaluation stage. Finally, simulation tests are performed on experimental data sets, and the residual series obtained from the output of this framework are examined for anomalous sensors using Bayesian estimation methods. The mean squared error (MSE), that is, the average squared difference between the reconstructed value and the actual standard value, is used as the evaluation index. The experimental results demonstrate that the MSE of the improved autoencoder is smaller than that of the unimproved model in both transient and steady-state conditions. In addition, the dynamic threshold method reduces the false alarm rate of anomaly detection.
The imbalanced data classification is a challenging issue in many domains including medical intelligent diagnosis and fraudulent transaction analysis. The performance of the conventional classifier degrades due to the...
详细信息
The imbalanced data classification is a challenging issue in many domains including medical intelligent diagnosis and fraudulent transaction analysis. The performance of the conventional classifier degrades due to the imbalanced class distribution of the training data set. Recently, machine learning and deep learning techniques are used for imbalanced data classification. Data preprocessing approaches are also suitable for handling class imbalance problem. Data augmentation is one of the preprocessing techniques used to handle skewed class distribution. Synthetic Minority Oversampling Technique (SMOTE) is a promising class balancing approach and it generates noise during the process of creation of synthetic samples. In this paper, autoencoder is used as a noise reduction technique and it reduces the noise generated by SMOTE. Further, Deep one-dimensional Convolutional Neural Network is used for classification. The performance of the proposed method is evaluated and compared with existing approaches using different metrics such as Precision, Recall, Accuracy, Area Under the Curve and Geometric Mean. Ten data sets with imbalance ratio ranging from 1.17 to 577.87 and data set size ranging from 303 to 284807 instances are used in the experiments. The different imbalanced data sets used are Heart-Disease, Mammography, Pima Indian diabetes, Adult, Oil-Spill, Phoneme, Creditcard, BankNoteAuthentication, Balance scale weight & distance database and Yeast data sets. The proposed method shows an accuracy of 96.1%, 96.5%, 87.7%, 87.3%, 95%, 92.4%, 98.4%, 86.1%, 94% and 95.9% respectively. The results suggest that this method outperforms other deep learning methods and machine learning methods with respect to G-mean and other performance metrics.
Graph clustering is a crucial technique for partitioning graph data. Recent research has concentrated on integrating topology and attribute information from attribute graphs to generate node embeddings, which are subs...
详细信息
Graph clustering is a crucial technique for partitioning graph data. Recent research has concentrated on integrating topology and attribute information from attribute graphs to generate node embeddings, which are subsequently clustered using classical algorithms. However, these methods have some limitations, such as insufficient information inheritance in shallow networks or inadequate quality of reconstructed nodes, leading to suboptimal clustering performance. To tackle these challenges, we introduce two normalization techniques within the graph attention autoencoder framework, coupled with an MSE loss, to facilitate node embedding learning. Furthermore, we integrate Transformers into the self-optimization module to refine node embeddings and clustering outcomes. Our model can induce appropriate node embeddings for graph clustering in a shallow network. Our experimental results demonstrate that our proposed approach outperforms the state-of-the-art in graph clustering over multiple benchmark datasets. In particular, we achieved 76.3% accuracy on the Pubmed dataset, an improvement of at least 7% compared to other methods.
This paper presents a novel topology optimization approach for the design of synchronous reluctance motors based on an autoencoder (AE) combined with the level set (LS) method. As the initial shapes of the LS method, ...
详细信息
This paper presents a novel topology optimization approach for the design of synchronous reluctance motors based on an autoencoder (AE) combined with the level set (LS) method. As the initial shapes of the LS method, the technique uses the shape generated by the AE, which learns the relationship between the objective function values and the design shapes in the optimization process. The proposed method trains the network parameters such that certain latent variable components represent shape features that are correlated with targeted objective functions. Consequently, shape variations that correspond to changes in multiple objective function values can be independently and continuously visualized. This enables the efficient preparation of new structures that are expected to have high performance. Finally, the AE-generated shapes are used as the initial shapes for LS optimization to derive practical Pareto solutions.
Metamodel-assisted optimization is a frequently applied approach for structural design optimization problems. Here, a data-driven metamodel approximates the computationally expensive simulation results of first princi...
详细信息
Metamodel-assisted optimization is a frequently applied approach for structural design optimization problems. Here, a data-driven metamodel approximates the computationally expensive simulation results of first principle models, e.g., finite element analyses. A significant drawback of typical metamodels is the limited amount of information that can be predicted due to their generally low-dimensional model output. Consequently, the metamodel usually does not predict the distribution of the desired quantity. This work presents a metamodel approach capable of predicting the spatial and temporal distribution of quantities for structural processes. This increases the modeling capability and makes more information available for the optimization. The autoencoder compresses the spatial distribution into a couple of features. The proposed methodology is applied to a three-stage forming process. Copyright (c) 2024 The Authors.
In this letter, we present a novel approach for denoising channel state information (CSI) feedback in massive multiple-input multiple-output (MIMO) cellular networks. Our method utilizes Deep Learning (DL) techniques ...
详细信息
In this letter, we present a novel approach for denoising channel state information (CSI) feedback in massive multiple-input multiple-output (MIMO) cellular networks. Our method utilizes Deep Learning (DL) techniques to compress and remove noise from measured CSI. Traditional DL-based denoising requires pairs of noisy input and corresponding clean targets, which are impractical to obtain in real-world wireless networks. To address this challenge, we propose a training method of denoising autoencoder using pairs of noisy CSIs and practical data acquisition strategies. Extensive evaluations demonstrate the superior reconstruction performance of our method compared to a vanilla autoencoder and legacy codebook-based CSI feedback.
Recently, researchers have leveraged the Denoising autoencoder (DAE) to reduce the noise in side-channel acquisitions (a.k.a. traces) that reduces the effectiveness of key recovery. Taking the ${L}2$ Loss (Mean Square...
详细信息
Recently, researchers have leveraged the Denoising autoencoder (DAE) to reduce the noise in side-channel acquisitions (a.k.a. traces) that reduces the effectiveness of key recovery. Taking the ${L}2$ Loss (Mean Square Error, MSE) as the objective function of the DAE, it only aims to lessen the Euclidean Distance (ED) between the input and output, overlooking the Intra-Data Correlation (IDC) of the trace which includes the timing information. This paper proposes the Multi-Loss Regularized Denoising autoencoder (ML-DAE) framework to improve the generalization capability of the DAE. This framework consists of a shared DAE and Multiple Loss (ML) functions that aim to reduce the noise while preserving the excellent IDC of the output. During the training phase, to avoid issues of overfitting and a high number of training parameters, we pre-train the DAE using MSE and then initiate the ML-DAE which contains a multicore Partial Loss (PL) function with parameters transferred from the pre-trained DAE. During the testing phase, the outputs from the multicore PL are fused using an average pooling layer to yield the final predictions. The experiments on highly noisy datasets (XMEGA_ME, DPA_V2, and AES_GPU) and the masked dataset ASCAD demonstrate that ML-DAE achieves an SNR gain of at least four times, hence Deep-Learning based Side-Channel Attacks (DLSCAs) and Template Attacks (TA) with denoising pre-processing reduce of the number of traces needed to recover the key in the attack phase by more than 55%.
暂无评论