Search

Role-based database management systems are most widely used for information storage and analysis but are known as vulnerable to insider attacks. The core of intrusion detection lies in an adaptive system, where an insider attack can be judged if it is different from the predicted role by performing classification on the user’s queries accessing the database and comparing it with the authorized role. In order to handle the high similarity of user queries for misclassified roles, this paper proposes a deep metric neural network with strategic sampling algorithm that properly extracts salient features and directly learns a quantitative measure of similarity. A strategic sampling method of heuristically generating and learning training pairs through Monte Carlo search is proposed to select a training pair that can represent the entire dataset. With the TPC-E–based benchmark data trained with 11,000 queries for 11 roles, the proposed model produces the classification accuracy of 95.41%, which is the highest compared with the previous models. The results are verified through comparison of quantitative and qualitative evaluations, and the feature space modelled in the neural network is analysed by t-SNE algorithm.

The prognostics and health management (PHM) aims to provide a comprehensive solution for equipment health care. Classifying the operation mode of excavator, one of the tasks in the PHM, is important to evaluate the remaining useful lifetime. Several studies have been conducted to classify the operations with either video or sensor data, but they have several limitations to use only one type of data. A model trained with sensor data cannot classify the similar operations such as “digging” and “ditch digging”, whereas a model with video data is vulnerable to surrounding condition like weather. In this paper, to overcome these shortcomings, we propose a deep neural network ensemble called FusionNet that classifies the operations of excavator. Two models are trained with sensor data and video frames respectively, where the feature extractors are transferred to the FusionNet. The proposed network ensemble performs a flexible and well-optimized classification by automatically calculating weights according to the extracted feature vectors and combining them. To verify the proposed model, several experiments are conducted with the real-world data. The proposed model achieves the accuracy of 99.17% which outperforms the conventional methods. We also confirm that the proposed model can address the shortcomings of using only one type of data and maximize the benefits through the automatic weighting of extracted features.

Database security focuses on protecting most organization’s virtual data storage unit and confidential information from malicious threats and external attacks. To keep out data secure, we need to use a role-based access control (RBAC) approach to accurately differentiate access permissions, but SQL queries written by an authorized user have very similar characteristics and are difficult to distinguish. In this paper, we propose a method of optimizing CNN-LSTM neural networks with particle swarm optimization (PSO) to classify the roles in RBAC system. Convolutional neural network (CNN) can extract parsed SQL queries into smaller details and features through an analysis mechanism. Long short-term memory (LSTM) is also suitable for modeling the temporal information of SQL queries to recognize the context of user authorities. PSO repeatedly searches and optimizes the complex hyperparameter space of the CNN-LSTM. Our PSO-based CNN-LSTM neural networks outperform other deep learning and machine learning models in the TPC-E benchmark SQL query statement. Finally, experiments and analysis show the usefulness of PSO and identify the important SQL query features that affect user role classification.

Noise from automobiles, such as buzzing, squeaking, and rattling (BSR) noises, is a key factor in automobile quality assessment. It is necessary to classify these noises for appropriate handling and prevention. Although many researchers have conducted studies to classify noise, they suffer from several problems: difficulty in extracting appropriate features, insufficient data to train a classifier, and weak robustness to surrounding noise. This paper proposes a method called latent semantic controlling generative adversarial networks (LSC-GAN) to solve these problems. To capture the features of data, a variational autoencoder (VAE), an autoencoder with approximate inference in a latent Gaussian model, learns the data representation by projecting them into the latent space according to their features and reconstructing the projected data. Because the generator and the discriminator of the LSC-GAN are trained simultaneously, the capacity to extract the characteristics of the data is improved and a knowledge space of classifiable data is also expanded with insufficient data. While data are generated by the generator, the encoder projects them back to the latent space according to their characteristics to advance the ability to extract features. Finally, the encoder is trained to the classifier, which is trained to classify BSR noises. The proposed classifier outperforms other models and achieves an accuracy of 96.68%. We confirm using a confusion matrix that the proposed model classifies the types of insufficient class better than other models. Our proposed model classifies data with accuracy of 94.68%, even if the data contains surrounding noise, which means it is more robust to BSR with surrounding noise than other models.

Photovoltaic (PV) panels are generally treated as the most dependable components of PV systems; therefore, investigations are necessary to understand and emphasize the degradation of PV cells. In almost all specific deprivation models, humidity and temperature are the two major factors that are responsible for PV module degradation. However, even if the degradation mode of a PV module is determined, it is challenging to research them in practice. Long-term response experiments should thus be conducted to investigate the influences of the incidence, rates of change, and different degradation methods of PV modules on energy production; such models can help avoid lengthy experiments to investigate the degradation of PV panels under actual working conditions. From the review, it was found that the degradation rate of PV modules in climates where the annual average ambient temperature remained low was -1.05% to -1.16% per year, and the degree of deterioration of PV modules in climates with high average annual ambient temperatures was -1.35% to -1.46% per year; however, PV manufacturers currently claim degradation rates of up to -0.5% per year.

Financial institutions, by and large, rely on the use of machine learning techniques to improve the classic credit risk assessment model for reduction of costs, delivery of faster decisions, guaranteed credit collections, and risk mitigations. As such, several data mining and machine learning approaches have been developed for computation of credit scores over the last few decades. Moreover, the existing rule‐based classification algorithms tend to generate a number of rules with a large number of conditions in the antecedent part. However, these algorithms fail to demonstrate high predictive accuracy while balancing coverage and simplicity. Thus, it becomes quite a challenging task for the researchers to generate an optimal rule set with high predictive accuracy. In this paper, we present an effective rule based classification technique for the prediction of credit risk using a novel Biogeography Based Optimization (BBO) method. The novel BBO in the context of rule mining is named as locally and globally tuned biogeography based rule‐miner (LGBBO‐RuleMiner). This is applied for discovering optimal rule set with high predictive accuracy from the dataset containing both the categorical and continuous attributes. The performance of the proposed algorithm is compared against a variety of rule‐miners such as OneR (1R), PART, JRip, Decision Table, Conjunctive Rule, J48, and Random Tree, along with some meta‐heuristic based rule mining techniques by considering two credit risk datasets obtained from University of California, Irvine (UCI) repository. It is found from the comparative study that the proposed rule miner in ten independent runs of ten‐fold cross validation outperforms all of the aforesaid algorithms in terms of predictive accuracy, coverage, and simplicity.

In recent industrial applications, machine learning technology is proving useful in preventing equipment failures in advance through early failure diagnosis. In particular, we show that different domains can be linked through adversarial learning with data available in different working conditions to facilitate the training of the model, as it is impractical to acquire data for all conditions in real-world applications. Nevertheless, the initial failure is a difficult problem to diagnose because it does not show a significant difference from the normal data between different conditions. Moreover, if only the domain discriminator is judged when adapting the domain, it tends to easily cause misclassification, so the reliability of the detection result needs to be improved. In this study, we propose a new learning method that improves classification performance by sharing the classification characteristics of the classifier for each task with the target domain characteristic generator. The proposed mechanism uses spatial attention to extract the focused partial information of the feature generator and discriminator, and further enhances task-specific features using the attention mechanism between the two extracted information types. Addresses the challenge of implementing both domain adaptation and classification. Extensive experimentation demonstrates efficiency and improved classification performance on benchmark and real-world application datasets. In real machine cases, the classification accuracy is improved by almost 4%. In addition, the negative impact on false alarms was lowered by increasing the classification accuracy of minimum failures. Convincingly demonstrate model effectiveness by performing an empirical analysis of the method through ablation analysis and visualization.

Prediction of currency exchange rate becomes highly desirable due to its greater role in financial and managerial decision making process. The fluctuations in exchange rate affect the economy of a country. Hence, over the years different types of neural network models along with statistical models are developed to predict the currency exchange rates of different countries with varying parameters. In this paper, we divide our effort into two parts. In first part, we have reviewed a few selected models of neural networks and statistics including fundamental and technical aspects of currency exchange rate prediction. Additionally, a thorough and careful experimental result analysis has been conducted on the models reviewed in part one. A committee machine has been proposed in part two to address the shortcomings of both neural networks and statistical models in the context of exchange rate prediction. Our study reveals that the currency exchange rates with multi-layer neural networks having Bayesian learning predictive accuracy is better than multi-layer neural networks with back-propagation learning. However, in the case of higher-order neural network multi-stage radial basis function network is predicting better than single stage radial basis function network. In the case of statistical models, it is drawn that under the umbrella of root mean square error measure, random walk is predicting better than other models of this category, whereas variance based model predicts better than rest of the models grouped under normalized mean square error measure. On the other hand, the integrated model is performing better than its counterpart like models with stand-alone mode. Moreover, our newly proposed committee machine is drawing a clear line over all the models while predicting exchange rate of GBP/USD.

Objectives: Uremic pruritus is common in patients with chronic kidney disease (CKD). The retention of uremic solutes is thought to be associated with uremic pruritus. Meanwhile, activation of protease-activated receptor-2 (PAR-2) has been suggested to play an important role in pruritus. The present study was performed to investigate the effects of uremic solutes on the expression of PAR-2 in the skin. Methods: Indoxyl sulfate (IS), p-cresol (PC), and uremic sera from CKD patients were used to stimulate PAR-2 expression in normal human epidermal keratinocytes (NHEKs). Also, NHEKs were additionally pretreated with soybean trypsin inhibitor to evaluate its inhibitory effect on PAR-2 expression. Patterns of cutaneous PAR-2 expression were investigated in skin samples from five CKD patients and CKD mice. Results: In NHEKs, IS, PC, and sera from CKD patients significantly induced PAR-2 mRNA and protein expression. Soybean trypsin inhibitor significantly decreased PAR-2 mRNA and protein expression in NHEKs treated with IS, PC, and CKD sera. NHEKs treated with IS and PC exhibited significant increases in protease activity. Skin from both CKD patients and mice exhibited marked upregulation of PAR-2 expression compared to control skin. Conclusions: Results from the present study suggest that uremic solutes either directly or indirectly affect PAR-2 expression in the skin of CKD subjects, potentially playing an important role in the pathogenesis of uremic pruritus.

This work presents a novel approach by considering teaching learning based optimization (TLBO) and radial basis function neural networks (RBFNs) for building a classifier for the databases with missing values and irrelevant features. The least square estimator and relief algorithm have been used for imputing the database and evaluating the relevance of features, respectively. The preprocessed dataset is used for developing a classifier based on TLBO trained RBFNs for generating a concise and meaningful description for each class that can be used to classify subsequent instances with no known class label. The method is evaluated extensively through a few bench-mark datasets obtained from UCI repository. The experimental results confirm that our approach can be a promising tool towards constructing a classifier from the databases with missing values and irrelevant attributes.

Theory of mind (ToM) is the ability to understand others’ mental states (e.g., intentions). Studies on human ToM show that the way we understand others’ mental states is very efficient, in the sense that observing only some portion of others’ behaviors can lead to successful performance. Recently, ToM has gained interest in robotics to build robots that can engage in complex social interactions. Although it has been shown that robots can infer others’ internal states, there has been limited focus on the data utilization of ToM mechanisms in robots. Here we show that robots can infer others’ intentions based on limited information by selectively and flexibly using behavioral cues similar to humans. To test such data utilization, we impaired certain parts of an actor robot’s behavioral information given to the observer, and compared the observer’s performance under each impairment condition. We found that although the observer’s performance was not perfect compared to when all information was available, it could infer the actor’s mind to a degree if the goal-relevant information was intact. These results demonstrate that, similar to humans, robots can learn to infer others’ mental states with limited information.

As communication technology develops, various sensory stimuli can be collected in service spaces. To enhance the service effectiveness, it is important to determine the optimal stimuli to induce group emotion in the service space to the target emotion. In this paper, we propose a stimuli control system to adjust the group emotion. It is a stand-alone system that can determine optimal stimuli by utility table and modular tree-structured Bayesian networks designed for emotion prediction model proposed in the previous study. To verify the proposed system, we collected data using several scenarios at a kindergarten and a senior welfare center. Each space is equipped with sensors for collection and equipment for controlling stimuli. As a result, the system shows a performance of 78% in the kindergarten and 80% in the senior welfare center. The proposed method shows much better performance than other classification methods with lower complexity. Also, reinforcement learning is applied to improving the accuracy of stimuli decision for a positive effect on system performance.

Role-based access control (RBAC) in databases provides a valuable level of abstraction to promote security administration at the business enterprise level. With the capacity for adaptation and learning, machine learning algorithms are suitable for modeling normal data access patterns based on large amounts of data and presenting robust statistical models that are not sensitive to user changes. We propose a convolutional neural-based learning classifier system (CN-LCS) that models the role of queries by combining conventional learning classifier system (LCS) with convolutional neural network (CNN) for a database intrusion detection system based on the RBAC mechanism. The combination of modified Pittsburgh-style LCSs for the optimization of feature selection rules and one-dimensional CNNs for modeling and classification in place of traditional rule generation outperforms other machine learning classifiers on a synthetic query dataset. In order to quantitatively compare the inclusion of rule generation and modeling processes in the CN-LCS, we have conducted 10-fold cross-validation tests and analysis through a paired sampled t-test.

Spanish Ministry of Science and Innovation [MINECO-TIN2017-84804-R, PID2020-112726RB-I00]; State Research Agency (AEI, Spain) [RED2018-102312-T]; Korean government (MSIT) (Yonsei University) [2020-0-01361]; AI Graduate School Program (Yonsei University)

Recommender systems are a long-standing research problem in data mining and machine learning. They are incremental in nature, as new user-item interaction logs arrive. In real-world applications, we need to periodically train a collaborative filtering algorithm to extract user/item embedding vectors and therefore, a time-series of embedding vectors can be naturally defined. We present a time-series forecasting-based upgrade kit (TimeKit), which works in the following way: it i) first decides a base collaborative filtering algorithm, ii) extracts user/item embedding vectors with the base algorithm from user-item interaction logs incrementally, e.g., every month, iii) trains our time-series forecasting model with the extracted time-series of embedding vectors, and then iv) forecasts the future embedding vectors and recommend with their dot-product scores owing to a recent breakthrough in processing complicated time-series data, i.e., neural controlled differential equations (NCDEs). Our experiments with four real-world benchmark datasets show that the proposed time-series forecasting-based upgrade kit can significantly enhance existing popular collaborative filtering algorithms., Comment: Accepted at IEEE BigData 2022

Aromatic compounds are organic matters that have the form of carbon rings, such as benzene rings. The compound is contained in most real-life chemicals, including medicines, detergents, and cosmetics. Because mutagenicity of these substances can pose a great health risk if they enter the human body, predicting mutagenicity and avoiding risk are important problems. Due to recent advances in deep learning, many studies have been conducted to predict the mutagenicity of graphically expressed aromatic compounds, and they are performing well. However, previous methods that used deep learning to predict the mutagenicity of molecules lead to the problem of dilution of local information in molecular structures. In graph neural networks, the embedding of graphs is determined by the average or sum of node embeddings, and when a particular node dominates, information from local nodes is diluted. In this paper, we propose a model that learns molecules’ local information properties to be undiluted by separating them from the original graph. By partitioning the graph, we preserve local information by breaking the relationship between the dominant nodes and the non-dominant nodes so that they do not affect each other's information updates. Experiments show that we successfully partition carbon rings and functional groups in molecular graphs using the Girvan Newman algorithm, and embed the segmented graphs using the same neural networks, allowing neural networks to learn about carbon rings and functional groups. Comparisons with other neural networks confirm performance improvements on 1% accuracy.

In this paper, we propose a novel method of cluster analysis called unsupervised functional link artificial neural networks (UFLANNs), which inherit the best characteristics of functional link artificial neural networks and self-organizing feature maps (SOFMs). UFLANNs adopt three types of basis functions such as Chebyshev, Legendre orthogonal polynomials, and power series for mapping the input data into a new feature space with higher dimensions, where the objects are clustered based on the principle of competitive learning of SOFMs. The effectiveness of this algorithm has been tested with various artificial and real-life datasets including remote sensing images. A thorough comparison with other popular clustering algorithms shows that the proposed method is promising in revealing clusters from many complex datasets.

Spanish Ministry of Economy and Competitiveness [AYA2014-57648-P]; regional Ministry of Economy and Employment, Asturias [FC-15-GRUPIN14-017]; Institute of Information & Communications Technology Planning & Evaluation (IITP) - Korean government (MSIT) [2020-0-01361]; 2019 IT promotion fund (Development of AI based Precision Medicine Emergency System) of the Korean government (MSIT), Jove, E., Blanco Rodríguez, P., Casteleiro Roca, J. L., Quintián, H., Moreno Arboleda, F. J., López Vázquez, J. A., Rodríguez Gómez, B. A., Meizoso López, M. C., Piñón Pazos, A., Cos Juez, F. J., Cho, S. B., Calvo Rolle, J. L.

Relational database management system (RDBMS) is the most popular database system. It is important to maintain data security from information leakage and data corruption. RDBMS can be attacked by an outsider or an insider. It is difficult to detect an insider attack because its patterns are constantly changing and evolving. In this paper, we propose an adaptive database intrusion detection system that can be resistant to potential insider misuse using evolutionary reinforcement learning, which combines reinforcement learning and evolutionary learning. The model consists of two neural networks, an evaluation network and an action network. The action network detects the intrusion, and the evaluation network provides feedback to the detection of the action network. Evolutionary learning is effective for dynamic patterns and atypical patterns, and reinforcement learning enables online learning. Experimental results show that the performance for detecting abnormal queries improves as the proposed model learns the intrusion adaptively using Transaction Processing performance Council-E scenario-based virtual query data. The proposed method achieves the highest performance at 94.86%, and we demonstrate the usefulness of the proposed method by performing 5-fold cross-validation.

Recently, the ear biometric has received much attention for human recognition due to its unique shape and rich local features. However, extracting discriminative features from ear images is a crucial task in presence of illumination changes, low contrast, noise, and pose variations. With the aim of neutralizing the effect of these factors, this paper proposes an automatic enhancement technique using meta-heuristic optimization to enhance the ear images. Here, we modified a recent and simple yet meta-heuristic optimization technique known as Jaya algorithm by introducing a mutation operator to enhance the ear images in few iterations and the proposed approach is named as enhanced Jaya algorithm. Then, we employed a pose-invariant local feature extractor, SURF to extract local features. Finally, the k-NN classifier has used to evaluate the rate of correct identification. Extensive experiments are conducted on four standard datasets and the performance evaluation is carried out by qualitative and quantitative measures. Experimental results clearly indicate the proposed enhancement approach is competitive as compared to two classical methods HE, CLAHE, and two meta-heuristic algorithms PSO and DE-based image enhancement techniques.

The rapid increase in human population and development in technology have sharply raised power consumption in today's world. Since electricity is consumed simultaneously as it is generated at the power plant, it is important to accurately predict the energy consumption in advance for stable power supply. In this paper, we propose a CNN-LSTM neural network that can extract spatial and temporal features to effectively predict the housing energy consumption. Experiments have shown that the CNN-LSTM neural network, which combines convolutional neural network (CNN) and long short-term memory (LSTM), can extract complex features of energy consumption. The CNN layer can extract the features between several variables affecting energy consumption, and the LSTM layer is appropriate for modeling temporal information of irregular trends in time series components. The proposed CNN-LSTM method achieves almost perfect prediction performance for electric energy consumption that was previously difficult to predict. Also, it records the smallest value of root mean square error compared to the conventional forecasting methods for the dataset on individual household power consumption. The empirical analysis of the variables confirms what affects to forecast the power consumption most.

Classification is a key factor in accuracy, simplicity, and expressiveness, and it is difficult to optimize all of these factors at the same time. The learning classifier system (LCS) is a suitable technique for addressing an adaptive classification problem. It is a combination of fast approximation and evolutionary optimization techniques. A neural-based learning classifier system (N-LCS) includes an architecture for maintaining expressiveness by incorporating neural networks into a supervised classifier system, which is also an LCS specializing in classification studies. In recent years, studies using deep artificial neural networks have been actively conducted. In particular, deep convolutional neural networks (CNN) provide a powerful representation in an extremely fundamental method and demonstrates the high performance in various domains. In this paper, we exploit various deep CNN architectures in convolutional neural-based learning classifier systems (CN-LCS) combining the CNN and LCS to explore the possibility of a CN-LCS. By using various CNNs as an action of a classifier in an N-LCS, better classification accuracy can be obtained and classifier can be optimized. Experimental results show that our models achieve the higher performance than N-LCS for database intrusion detection as well as two other datasets, and extract effective features from deep representation by projecting data samples learned by several deep CNN models into the feature space.

Soft keyboard has been popular since the smartphone made an appearance. As the smartphone is a personal device used for user's preference, we need to personalize the soft keyboard. In order to make a personal context-aware keyboard, a large amount of data is required from each individual, but it is difficult to collect the data enough. In this paper, we propose a novel method to construct random forests with additional data of the same clusters, which copes with small amount of the individual data by adding similar data through K-means clustering algorithm. In addition, for the personalization of the soft keyboard, the preferred GUIs are recommended according to the activities and input hand postures of the user recognized by the random forest models trained with the clustered data. To train the proposed system, we have collected the data from 200 people. Each person can use the most necessary keyboard by selecting the appropriate GUI for each situation depending on the smartphone usage activity and input hand postures. The proposed system showed better performance than a model using all common data and a model using individual data only.

Social lending is made between peers, and with the risk that the investor can take direct damages from the borrower’s failure to repay, accurate default prediction for borrowers is important. The repayment result can be known after the end of the repayment period, and such data is limited. However, social loans are matched online in real time and large amounts of unlabeled data are being generated. In this paper, we propose a method to combine label propagation and transductive support vector machine (TSVM) with Dempster–Shafer theory for accurate default prediction of social lending using unlabeled data. In order to train a lot of data effectively, we ensemble semi-supervised learning methods with different characteristics. Label propagation is performed so that data having similar features are assigned to the same class and TSVM makes moving away data having different features. Dempster–Shafer fusion method allows accurate labeling by exploiting the merits of the two methods. Experiments are performed using the open data set from Lending Club. The accuracy of the proposed method is improved by about 10% against that of the model using only labeled data, and more accurate labeling can be performed through the proposed ensemble method.

Considering the fatality of phishing attacks that are emphasized by many organizations, the inductive learning approach using reported malicious URLs has been verified in the field of deep learning. However, the deep learning-based method mainly focused on the fitting of a classification task via historical URL observation shows a limitation of recall due to the characteristics of zero-day attack. In order to model the nature of a zero-day phishing attack in which URL addresses are generated and discarded immediately, an approach that utilizes the expert knowledge is promising. We introduce the integration method of deep learning and logic programmed domain knowledge to inject the real-world constraints. We design neural and logic classifiers and propose the joint learning method of each component based on the traditional neuro-symbolic integration. Extensive experiments on three real-world datasets consisting of 222,541 URLs showed the highest recall among the latest deep learning methods, despite the hostile class-imbalanced condition. We demonstrate that the optimized weighting between neural and logic component has an effect of improving the recall over 3% compared to the existing methods.