Neural networks are well-known for their powerful capability in producing high prediction accuracy. However, due to the non-linear calculations in the network, it is very difficult for users to understand which input ...
详细信息
ISBN:
(纸本)9781728104577
Neural networks are well-known for their powerful capability in producing high prediction accuracy. However, due to the non-linear calculations in the network, it is very difficult for users to understand which input features are important in leading to final predictions. In this study, we propose a two-step pipeline approach that uses two sets of linear models to estimates feature importance in the input dataset X that leads to the class prediction specified in Y. More specifically, the first linear regression model derives the feature importance in X in explaining the Z-code that was extracted from any hidden layer of a trained neural network. The second linear classification model captures the importance in the Z-code in predicting the target class Y. We then combine the first X to Z importance with the second Z to Y importance together to approximate the non-linear importance from X to Y. The experiments conducted in this study also show that our method is sound and stable in selecting the truly important features from input datasets regardless how a neural network was constructed with different parameters such as activation functions or the number of hidden layers.
In many visual domains like fashion, building an effective unsupervised clustering model depends on visual feature representation instead of structured and semi-structured data. In this paper, we propose a fashion ima...
详细信息
ISBN:
(纸本)9783030214517;9783030214500
In many visual domains like fashion, building an effective unsupervised clustering model depends on visual feature representation instead of structured and semi-structured data. In this paper, we propose a fashion image deep clustering (FiDC) model which includes two parts, feature representation and clustering. The fashion images are used as the input and are processed by a deep stacked autoencoder to produce latent feature representation, and the output of this autoencoder will be used as the input of the clustering task. Since the output of the former has a great influence on the later, the strategy adopted in the model is to integrate the learning process of the autoencoder and the clustering together. The autoencoder is trained with the optimal number of neurons per hidden layers to avoid overfitting and we optimize the cluster centroid by using stochastic gradient descent and backpropagation algorithm. We evaluate FiDC model on a real-world fashion dataset downloaded from Amazon where images have been extracted into 4096-dimensional visual feature vectors by convolutional neural networks. The experimental results show that our model achieves state-of-the-art performance.
The purpose of this paper is to study data fusion applications in traditional, spatial, and aerial video stream applications which addresses the processing of data from multiple sources using co-occurrence information...
详细信息
ISBN:
(纸本)9781510626584
The purpose of this paper is to study data fusion applications in traditional, spatial, and aerial video stream applications which addresses the processing of data from multiple sources using co-occurrence information and uses a common semantic metric. Use of co-occurrence information to infer semantic relations between measurements avoids the need to make use of such external information, such as labels. Many of the current Vector Space Models (VSM) do not preserve the co-occurrence information, leading to a less than useful similarity metric. We propose a proximity matrix embedding part of the learning metric representation which has entries showing the relations between co-occurrence frequency observed in input sets. First, we show an implicit spatial sensor proximity matrix calculation using Jaccard similarity for an array of sensor measurements and compare with the state-of-the-art kernel PCA learning from feature space proximity representation;it relates to a k-radius ball of nearest neighbors. Finally, we extend the class co-occurrence boosting of our unsupervised model using pre-trained multi-modal reuse.
In this paper, we investigate the cell outage detection in Self-Organizing Networks. The purpose of cell outage detection is to automatically detect whether there exist some failures or degradation in the base station...
详细信息
ISBN:
(纸本)9781728132488
In this paper, we investigate the cell outage detection in Self-Organizing Networks. The purpose of cell outage detection is to automatically detect whether there exist some failures or degradation in the base stations, such that users could not obtain mobile services, or the obtained mobile services do not fulfill their requirements. The cell outage detection in 5G is with great challenge. The deployment of future 5G mobile communication networks would be heterogeneous and ultra-dense. The mobile communication environments are very complicated. They include the multipath transmission, fading, shadowing, interference, and so on. Users' mobility and usage pattern also vary. In such environments, the mobile data would be large-scale and high-dimensional. Traditional small-scale and low-dimensional anomaly detection methods would be unsuitable. Moreover, operational mobile communication networks should be normal almost all the time. Cell outage would be seldom. Therefore, the normal data and anomaly data would be imbalanced. In this paper, we formulate the cell outage detection problem as an anomaly detection problem. We propose an cell outage detection method using the autoencoder, which is a neural network that is trained by unsupervised learning. The network could be trained in advance even when the cell outage data is still not available. Moreover, the autoencoder is also useful for denoising. This proposed method could thus automatically detect the cell outage in complicated and time-varying mobile wireless communication environments. Comprehensive system-level simulations validate the performance of the proposed method.
The development of object detection systems is normally driven to achieve both high detection and low false positive rates in a certain public dataset. However, when put into a real scenario the result is generally an...
详细信息
ISBN:
(纸本)9783030313210;9783030313203
The development of object detection systems is normally driven to achieve both high detection and low false positive rates in a certain public dataset. However, when put into a real scenario the result is generally an unacceptable rate of false alarms. In this context we propose to add an additional step that models and filters the typical false alarms of the new scenario while roughly maintaining the ability to detect the objects of interest. We propose to use the false alarms of the new scenario to train a deep autoencoder and to model them. The latter will act as a filter that checks whether the output of the detector is one of its typical false positives or not based on the reconstruction error measured with the Mean Squared Error (MSE) and the Peak Signal-to-Noise Ratio (PSNR). We test the system using an entirely synthetic novel dataset for training and testing the autoencoder generated with Unreal Engine 4. Results show a reduction in the number of FPs of up to 37.9% in combination with the PSNR error while maintaining the same detection capability.
Pathology reports are a main source of data for cancer surveillance programs. Manual coding of pathology reports is labor-intensive but necessary for obtaining labeled data to train automated information extraction sy...
详细信息
ISBN:
(纸本)9781728108483
Pathology reports are a main source of data for cancer surveillance programs. Manual coding of pathology reports is labor-intensive but necessary for obtaining labeled data to train automated information extraction systems. In this study, we investigated semi-supervised deep learning, improving the performance of a multitask information extraction system for automated annotation of pathology reports. We used a set of over 374,000 pathology reports from the Louisiana Tumor Registry and a novel convolutional attention-based auto-encoder. We performed a set of experiments comparing supervised training augmented with unlabeled data at 1%, 5%, 10%, and 50% of the original data size. We also compared the impact of extending text processing to include unlabeled tokens. We find that semi-supervised training consistently improved individual performance with increased micro-averaged F-scores between 0.012 and 0.064 and increased macro-averaged F-scores of up to 0.158. This demonstrates that semantic information learned via unsupervised learning can be used to improve supervised clinical task performance.
We present a latent space factorization that controls a generative neural network for shapes in a semantic way. Our method uses the segmentation data present in a collection of shapes to explicitly factorize the encod...
详细信息
ISBN:
(纸本)9783030269807;9783030269791
We present a latent space factorization that controls a generative neural network for shapes in a semantic way. Our method uses the segmentation data present in a collection of shapes to explicitly factorize the encoder of a pointcloud autoencoder network, replacing it by several sub-encoders. This allows to learn a semantically-structured latent space in which we can uncover statistical modes corresponding to semantically similar shapes, as well as mixing parts from several objects to create hybrids and quickly explore design ideas through varying shape combinations. Our work differs from existing methods in two ways: first, it proves the usefulness of neural networks to achieve shape combinations and second, adapts the whole geometry of the object to accommodate for its different parts.
We present two fully unsupervised deep learning approaches for hyperspectral anomaly detection. In one approach we formulate the anomaly detection problem as an adversarial game where a generator network learns the di...
详细信息
ISBN:
(纸本)9781728152943
We present two fully unsupervised deep learning approaches for hyperspectral anomaly detection. In one approach we formulate the anomaly detection problem as an adversarial game where a generator network learns the distribution of the hyperspectral background pixels comprising a single hyperspectral image and the output of the corresponding discriminator network yields a detection statistic. The other approach formulates the detection statistic as the error between an input hyperspectral pixel and the reconstruction of that pixel by an autoencoder network trained on the image. Both methods leverage a sub-sampling scheme that allows for unsupervised training and testing on the same data set. Our approaches are validated on a four-class synthetic hyperspectral data set and compared to a statistical approach (RX) and a geometric approach (skelton kernel principal component analysis). The proposed Generative Anomaly Detector algorithm achieves top performance on the data set while the autoencoder detection scheme also demonstrates performance gains relative to the comparison algorithms. Benefits and drawbacks of the approaches are discussed and highlight the many potential directions for future work.
Light field digital images are novel image modalities for capturing a sampled representation of the plenoptic function. A large amount of data is typically associated to a single sample of a scene, and data compressio...
详细信息
ISBN:
(纸本)9781728144962
Light field digital images are novel image modalities for capturing a sampled representation of the plenoptic function. A large amount of data is typically associated to a single sample of a scene, and data compression tools are required in order to develop systems and applications for light field communications. This paper presents the study of the performance of a convolutional neural network autoencoder as a tool for digital light field image compression. Testing conditions and a framework for the experimental evaluation are proposed for this study. Different encoders and coding conditions are taken into consideration, obtained results are reported and critically discussed.
暂无评论