Nowadays, machine learning is being used widely. There have also been attacks towards machine learning process. In this study, robustness against machine learning model attacks which cause many results such as misclas...
详细信息
ISBN:
(纸本)9781728119045
Nowadays, machine learning is being used widely. There have also been attacks towards machine learning process. In this study, robustness against machine learning model attacks which cause many results such as misclassification, disruption of decision mechanisms and avoidance of filters has been shown by autoencoding and with non-targeted attacks to a model trained with Mnist dataset. In this work, the results and improvements for the most common and important attack method, non-targeted attack are presented.
The paper proposes a neural model for a direct comparison of the two so-called Double Dummy Bridge Problem (DDBP) instances, along with a practical use-case for determining which pair, NS or WE, should propose the hig...
详细信息
ISBN:
(纸本)9781728119854
The paper proposes a neural model for a direct comparison of the two so-called Double Dummy Bridge Problem (DDBP) instances, along with a practical use-case for determining which pair, NS or WE, should propose the higher deal during a bidding phase in a Bridge game. The proposed system is composed of two identical subnetworks combined by a comparator layer placed on top of them. The base of each subnetwork is a shallow autoencoder (AE) which is further connected with a Multilayer Perceptron. The system is trained in two phases - an unsupervised one - used to create a meaningful feature-based input representation in AE compression layer, and a supervised one - meant for fine-tuning of the whole model. Training and test data are composed of pairs of Bridge deals in which the second deal in a pair is the first one rotated by 90 degrees. Since the task is to point which of the two deals promise a higher contract for the NS pair, due to deal rotation within a pair, the system effectively answers the title question "Who should bid higher, NS or WE, in a given deal?". The proposed approach is experimentally compared with two other methods: one relying on a neural system solving the DDBP and the other one employing several estimators of hand strength used by experienced players. The results clearly indicate that both neural network approaches outperform the usage of human-scoring systems by a large margin, most notably in the trump (suit) contract.
Background: Single-cell RNA sequencing (scRNA-seq) unfolds complex transcriptomic datasets into detailed cellular maps. Despite recent success, there is a pressing need for specialized methods tailored towards the fun...
详细信息
Background: Single-cell RNA sequencing (scRNA-seq) unfolds complex transcriptomic datasets into detailed cellular maps. Despite recent success, there is a pressing need for specialized methods tailored towards the functional interpretation of these cellular maps. Findings: Here, we present DrivAER, a machine learning approach for the identification of driving transcriptional programs using autoencoder-based relevance scores. DrivAER scores annotated gene sets on the basis of their relevance to user-specified outcomes such as pseudotemporal ordering or disease status. DrivAER iteratively evaluates the information content of each gene set with respect to the outcome variable using autoencoders. We benchmark our method using extensive simulation analysis as well as comparison to existing methods for functional interpretation of scRNA-seq data. Furthermore, we demonstrate that DrivAER extracts key pathways and transcription factors that regulate complex biological processes from scRNA-seq data. Conclusions: By quantifying the relevance of annotated gene sets with respect to specified outcome variables, DrivAER greatly enhances our ability to understand the underlying molecular mechanisms.
Multimodal emotion recognition is important for facilitating efficient interaction between humans and machines. To better detect emotional states from multimodal data, we need to effectively extract both the common in...
详细信息
ISBN:
(纸本)9781538695524
Multimodal emotion recognition is important for facilitating efficient interaction between humans and machines. To better detect emotional states from multimodal data, we need to effectively extract both the common information that captures dependencies among different modalities, and the private information that characterizes variations in each modality. However, existing works are mostly designed to pursue either one of these objectives but not both. In our work, we propose an end-to-end learning approach to simultaneously extract the common and private information for multimodal emotion recognition. Specifically, we use a correlation loss based on Hirschfeld-Gebelein-Renyi (HGR) maximal correlation and a reconstruction loss based on autoencoders to preserve the common and private information, respectively. Experimental results on eNTERFACE'05 database and RML database demonstrate the effectiveness of our proposed approach.
In this paper,we propose an image classification technique which uses a simple autoencoder with a regularizer. Nowadays, Convolutional Neural Networks (CNN) are primarily used for image classification. Our method can ...
详细信息
ISBN:
(纸本)9781728121536
In this paper,we propose an image classification technique which uses a simple autoencoder with a regularizer. Nowadays, Convolutional Neural Networks (CNN) are primarily used for image classification. Our method can be used for image classification with much reduced requirement of computational capability than a complex CNN which has a huge number of degrees of freedom. Here, the terms simple and complex, respectively, correspond to the simplicity and the complexity of a network in terms of the number of learnable parameters (degrees of freedom) and the number of hidden layers. This technique uses features extracted from a pretrained CNN, trained on a completely different dataset. Genetic algorithm solves for the optimal hyperparameters of the pretrained CNN. It is observed that these features serve as important and robust parameters for the training of the autoencoder, as a final average image classification accuracy improvement of about 17.45% is observed with the inclusion of these features. We use a pretrained CNN on MNIST dataset and classify images of several other benchmark datasets. We utilize different classifiers for image classification based on features extracted from the autoencoder and repeat each of the experiments a number of times with different random initialization of the classifier and the weight matrix of the autoencoder. We also perform experiments by pretraining the CNN with different datasets. Our results show a notable image classification accuracy and a significant reduction of training time with respect to a complex CNN.
User preferences are influenced by the purchased products, and ratings of products are also related to theirs public praises. Dynamic latent representations can be learned from these sequence information. Researches s...
详细信息
ISBN:
(纸本)9781728109459
User preferences are influenced by the purchased products, and ratings of products are also related to theirs public praises. Dynamic latent representations can be learned from these sequence information. Researches show that learning such dynamic features is helpful to build model-based collaborative filtering. However, static features also play an irreplaceable role in recommendations by reason of inherent characteristics of users/items. Ratings of users on products directly represent user preferences and qualities of products. A neural network model for learning both static and dynamic features is proposed in this paper. autoencoder is adopted as a static model focusing on explicit feedback i.e. ratings, and gated recurrent unit is adopted as a dynamic model focusing on implicit feedback i.e. sequences. Features learned from static and dynamic models are combined to make predictions. Experiments on two real-word datasets i.e. Baby of Amazon dataset and MovieLens 10M show improvement of our proposed model over the state-of-the-art methods.
Spectral clustering is one of the most popular modern clustering algorithms. It is easy to implement, can be solved efficiently, and very often outperforms other traditional clustering algorithms such as k-means. Howe...
详细信息
ISBN:
(纸本)9781728146034
Spectral clustering is one of the most popular modern clustering algorithms. It is easy to implement, can be solved efficiently, and very often outperforms other traditional clustering algorithms such as k-means. However, spectral clustering would be insufficient when dealing with most datasets which have complex statistical properties and requires the user to specify the number of clusters (called k). To address these two problems, in this paper, we propose an approach to extending spectral clustering with deep embedding and estimation of the number of clusters. Specifically, we first generate the deep embedding via learning a deep autoencoder, which transforms the raw data into the lower dimensional representations that suitable for clustering. We then provide an effective method to estimate the number of clusters by learning a softmax autoencoder from the deep embedding. We finally extend spectral clustering with the learned embedding and the estimated number. An extensive experimental study on several image and text datasets illustrates the effectiveness and efficiency of our approach.
Nowadays, more and more massive cyber-attacks have been launched over social networks. Using compromised or fake accounts, criminals can exploit the inherent trust between connected users to effectively spread malicio...
详细信息
ISBN:
(纸本)9781728135557
Nowadays, more and more massive cyber-attacks have been launched over social networks. Using compromised or fake accounts, criminals can exploit the inherent trust between connected users to effectively spread malicious content and perform scams against users. Detecting those malicious accounts on social networks like Twitter has received increasing attention from government, industry, and academia. Traditional methods on malicious account detection often leverage features that are created and selected based on domain knowledge of user data, which is inefficient, time-consuming, and possibly biased due to different understanding and observations of the data. In this paper, we propose a new framework, called MADAFE, for accurate and efficient malicious account detection on social networks like Twitter. To overcome the limitation of existing work on manual feature extraction, MADAFE utilizes an autoencoder to automate feature extraction and selection from unlabeled user data. A softmax regression model is also established and trained with the extracted features for classification of benign and malicious accounts. We test MADAFE on different datasets, and extensive simulation results show that MADAFE is effective in detecting malicious accounts, which outperforms state-of-the-art detection methods.
Frauds are known to be dynamic and have no patterns, hence they are not easy to identify. Fraudsters use recent technological advancements to their advantage. They somehow bypass security checks, leading to the loss o...
详细信息
ISBN:
(纸本)9781728137780
Frauds are known to be dynamic and have no patterns, hence they are not easy to identify. Fraudsters use recent technological advancements to their advantage. They somehow bypass security checks, leading to the loss of millions of dollars. Analyzing and detecting unusual activities using data mining techniques is one way of tracing fraudulent transactions. transactions. This paper aims to benchmark multiple machine learning methods such as k-nearest neighbor (KNN), random forest and support vector machines (SVM), while the deep learning methods such as autoencoders, convolutional neural networks (CNN), restricted boltzmann machine (RBM) and deep belief networks (DBN). The datasets which will be used are the European (EU) Australian and German dataset. The Area Under the ROC Curve (AUC), Matthews Correlation Coefficient (MCC) and Cost of failure are the 3-evaluation metrics that would be used.
In recent years, with the widespread use of encrypted traffic communication technology, network traffic encryption has been gradually becoming a standard of communication. This phenomenon has a great impact on traditi...
详细信息
ISBN:
(纸本)9781728125428
In recent years, with the widespread use of encrypted traffic communication technology, network traffic encryption has been gradually becoming a standard of communication. This phenomenon has a great impact on traditional traffic detection methods, especially on anomaly detection methods, which are highly dependent on the type of network protocol types and traffic data. By surveying the existing encrypted traffic classification and analyzing these methods, we found that there are two main methods for detection: payload-based detection and feature-based detection. On this basis, this paper puts forward an Encrypted Malicious Traffic Detection System which is based on multi-AEs(autoencoder). We use malicious sandbox for traffic data collection, mark malicious flow and normal flow with labels. After that, we use multilayer networks of AEs for feature extraction and training classifier model. Our system analyzes the feature of cryptographic protocol from handshake phase to Authentication phase on the basis of existing research, and we extract the traffic feature for a better classification by further expanding flow feature vector to the higher dimensions. In addition, we compared the performance of our model with the traditional learning model under the same environment. The experimental results showed that our system had higher detection accuracy and lower loss rate. We also studied the influence of dataset imbalance on results in the detection of encrypted traffic by several experiments. According to the experimental evaluation, dataset imbalance will lead to the decrease of positive rate.
暂无评论