TCAD simulation was performed to investigate the material properties of an AlGaN/GaN structure in Deep Acceptor (DA)-rich and Deep Donor (DD)-rich GaN cases. DD-rich semi-insulating GaN generated a positively charged area thereof to prevent the electron concentration in 2DEG from decreasing, while a DA-rich counterpart caused electron depletion, which was the origin of the current collapse in AlGaN/GaN HFETs. These simulation results were well verified experimentally using three nitride samples including buffer-GaN layers with carbon concentration ([C]) of 5×1017, 5×1018, and 4×1019 cm-3. DD-rich behaviors were observed for the sample with [C]=4×1019 cm-3, and DD energy level EDD=0.6 eV was estimated by the Arrhenius plot of temperature-dependent IDS. This EDD value coincided with the previously estimated EDD. The backgate experiments revealed that these DD-rich semi-insulating GaN suppressed both current collapse and buffer leakage, thus providing characteristics desirable for practical usage.

Double modular redundancy (DMR) is to execute operations twice and detect soft error by comparing the operation results. The error is corrected by executing necessary operations again. For the DMR design of conditional processing, a method is proposed which makes the secondary executions of the duplicated operations be dependent on the primary execution of the condition operation, thereby widening the schedule solution space and allowing better results to be derived. The energy minimization with the proposed method is formulated as ILP models and the optimum solution is obtained by using an ILP solver.

Most existing methods of effort estimations in software development are manual, labor-intensive and subjective, resulting in overestimation with bidding fail, and underestimation with money loss. This paper investigates effectiveness of sequence models on estimating development effort, in the form of man-months, from software project data. Four architectures; (1) Average word-vector with Multi-layer Perceptron (MLP), (2) Average word-vector with Support Vector Regression (SVR), (3) Gated Recurrent Unit (GRU) sequence model, and (4) Long short-term memory (LSTM) sequence model are compared in terms of man-months difference. The approach is evaluated using two datasets; ISEM (1,573 English software project descriptions) and ISBSG (9,100 software projects data), where the former is a raw text and the latter is a structured data table explained the characteristic of a software project. The LSTM sequence model achieves the lowest and the second lowest mean absolute errors, which are 0.705 and 14.077 man-months for ISEM and ISBSG datasets respectively. The MLP model achieves the lowest mean absolute errors which is 14.069 for ISBSG datasets.

Objective interestingness measures play a vital role in association rule mining of a large-scaled database because they are used for extracting, filtering, and ranking the patterns. In the past, several measures have been proposed but their similarities or relations are not sufficiently explored. This work investigates sixty-one objective interestingness measures on the pattern of A → B, to analyze their similarity and dissimilarity as well as their relationship. Three-probability patterns, P(A), P(B), and P(AB), are enumerated in both linear and exponential scales and each measure's values of those conditions are calculated, forming synthesis data for investigation. The behavior of each measure is explored by pairwise comparison based on these three-probability patterns. The relationship among the sixty-one interestingness measures has been characterized with correlation analysis and association rule mining. In the experiment, relationships are summarized using heat-map and association rule mined. As the result, selection of an appropriate interestingness measure can be realized using the generated heat-map and association rules.

The triaxial accelerometer is one of the most important sensors for human activity recognition (HAR). It has been observed that the relations between the axes of a triaxial accelerometer plays a significant role in improving the accuracy of activity recognition. However, the existing research rarely focuses on these relations, but rather on the fusion of multiple sensors. In this paper, we propose a data fusion-based convolutional neural network (CNN) approach to effectively use the relations between the axes. We design a single-channel data fusion method and multichannel data fusion method in consideration of the diversified formats of sensor data. After obtaining the fused data, a CNN is used to extract the features and perform classification. The experiments show that the proposed approach has an advantage over the CNN in accuracy. Moreover, the single-channel model achieves an accuracy of 98.83% with the WISDM dataset, which is higher than that of state-of-the-art methods.

Malicious attackers on the Internet use automated attack programs to disrupt the use of services via mass spamming, unnecessary bulletin boarding, and account creation. Completely automated public turing test to tell computers and humans apart (CAPTCHA) is used as a security solution to prevent such automated attacks. CAPTCHA is a system that determines whether the user is a machine or a person by providing distorted letters, voices, and images that only humans can understand. However, new attack techniques such as optical character recognition (OCR) and deep neural networks (DNN) have been used to bypass CAPTCHA. In this paper, we propose a method to generate CAPTCHA images by using the fast-gradient sign method (FGSM), iterative FGSM (I-FGSM), and the DeepFool method. We used the CAPTCHA image provided by python as the dataset and Tensorflow as the machine learning library. The experimental results show that the CAPTCHA image generated via FGSM, I-FGSM, and DeepFool methods exhibits a 0% recognition rate with ε=0.15 for FGSM, a 0% recognition rate with α=0.1 with 50 iterations for I-FGSM, and a 45% recognition rate with 150 iterations for the DeepFool method.

Since deep learning was introduced, a series of achievements has been published in the field of automatic machine translation (MT). However, Korean-Vietnamese MT systems face many challenges because of a lack of data, multiple meanings of individual words, and grammatical diversity that depends on context. Therefore, the quality of Korean-Vietnamese MT systems is still sub-optimal. This paper discusses a method for applying Named Entity Recognition (NER) and Part-of-Speech (POS) tagging to Vietnamese sentences to improve the performance of Korean-Vietnamese MT systems. In terms of implementation, we used a tool to tag NER and POS in Vietnamese sentences. In addition, we had access to a Korean-Vietnamese parallel corpus with more than 450K paired sentences from our previous research paper. The experimental results indicate that tagging NER and POS in Vietnamese sentences can improve the quality of Korean-Vietnamese Neural MT (NMT) in terms of the Bi-Lingual Evaluation Understudy (BLEU) and Translation Error Rate (TER) score. On average, our MT system improved by 1.21 BLEU points or 2.33 TER scores after applying both NER and POS tagging to the Vietnamese corpus. Due to the structural features of language, the MT systems in the Korean to Vietnamese direction always give better BLEU and TER results than translation machines in the reverse direction.

We propose an effective 2d image based end-to-end deep learning model for malware detection by introducing a black & white embedding to reserve bit information and adapting the convolution architecture. Experimental results show that our proposed scheme can achieve superior performance in both of training and testing data sets compared to well-known image recognition deep learning models (VGG and ResNet).

We propose a multi-targeted backdoor that misleads different models to different classes. The method trains multiple models with data that include specific triggers that will be misclassified by different models into different classes. For example, an attacker can use a single multi-targeted backdoor sample to make model A recognize it as a stop sign, model B as a left-turn sign, model C as a right-turn sign, and model D as a U-turn sign. We used MNIST and Fashion-MNIST as experimental datasets and Tensorflow as a machine learning library. Experimental results show that the proposed method with a trigger can cause misclassification as different classes by different models with a 100% attack success rate on MNIST and Fashion-MNIST while maintaining the 97.18% and 91.1% accuracy, respectively, on data without a trigger.

This article introduces our investigation on learning state estimation in e-learning on the condition that visual observation and recording of a learner's behaviors is possible. In this research, we examined methods of adaptation for a new learner for whom a small number of ground truth data can be obtained.

Internal user threats such as information leakage or system destruction can cause significant damage to the organization, however it is very difficult to prevent or detect this attack in advance. In this paper, we propose an anomaly-based insider threat detection method with local features and global statistics over the assumption that a user shows different patterns from regular behaviors during harmful actions. We experimentally show that our detection mechanism can achieve superior performance compared to the state of the art approaches for CMU CERT dataset.

This letter reveals that an edge-triggered master-slave flip-flop (FF) using well-known soft error tolerant DICE (dual interlocked storage cell) is vulnerable to soft errors occurring around clock edge. This letter presents a design of a soft error tolerant FF based on the master-slave FF using DICE. The proposed design modifies the connection between the master and slave latches to make the FF not vulnerable to these errors. The hardware overhead is almost the same as that for the original edge-triggered FF using the DICE.

In this paper, we propose a salient region detection method with multi-feature fusion and edge constraint. First, an image feature extraction and fusion network based on dense connection structure and multi-channel convolution channel is designed. Then, a multi-scale atrous convolution block is applied to enlarge reception field. Finally, to increase accuracy, a combined loss function including classified loss and edge loss is built for multi-task training. Experimental results verify the effectiveness of the proposed method.

In this paper, we propose a method to reduce the labeling cost while acquiring training data for a malicious domain name detection system using supervised machine learning. In the conventional systems, to train a classifier with high classification accuracy, large quantities of benign and malicious domain names need to be prepared as training data. In general, malicious domain names are observed less frequently than benign domain names. Therefore, it is difficult to acquire a large number of malicious domain names without a dedicated labeling method. We propose a method based on active learning that labels data around the decision boundary of classification, i.e., in the gray area, and we show that the classification accuracy can be improved by using approximately 1% of the training data used by the conventional systems. Another disadvantage of the conventional system is that if the classifier is trained with a small amount of training data, its generalization ability cannot be guaranteed. We propose a method based on ensemble learning that integrates multiple classifiers, and we show that the classification accuracy can be stabilized and improved. The combination of the two methods proposed here allows us to develop a new system for malicious domain name detection with high classification accuracy and generalization ability by labeling a small amount of training data.

End-to-end delay, aiming to realize how much time it will take for a traffic load generated by a Mobile Node (MN) to reach Sink Node (SN), is a principal objective of most new trends in a Wireless Sensor Network (WSN). It has a direct link towards understanding the minimum time delay expected where the packet sent by MN can take to be received by SN. Most importantly, knowing the average minimum transmission time limit is a crucial piece of information in determining the future output of the network and the kind of technologies implemented. In this paper, we take network load and transmission delay issues into account in estimating the Average Minimum Time Limit (AMTL) needed for a health operating cognitive WSN. To further estimate the AMTL based on network load, an end-to-end delay analysis mechanism is presented and considers the total delay (service, queue, ACK, and MAC). This work is proposed to answer the AMTL needed before implementing any cognitive based WSN algorithms. Various time intervals and cogitative channel usage with different application payload are used for the result analysis. Through extensive simulations, our mechanism is able to identify the average time intervals needed depending on the load and MN broadcast interval in any cognitive WSN.

Depression is a major mental health problem in Thailand. The depression rates have been rapidly increasing. Over 1.17 million Thai people suffer from this mental illness. It is important that a reliable depression screening tool is made available so that depression could be early detected. Given Facebook is the most popular social network platform in Thailand, it could be a large-scale resource to develop a depression detection tool. This research employs techniques to develop a depression detection algorithm for the Thai language on Facebook where people use it as a tool for sharing opinions, feelings, and life events. To establish the reliable result, Thai Mental Health Questionnaire (TMHQ), a standardized psychological inventory that measures major mental health problems including depression. Depression scale of the TMHQ comprises of 20 items, is used as the baseline for concluding the result. Furthermore, this study also aims to do factor analysis and reduce the number of depression items. Data was collected from over 600 Facebook users. Descriptive statistics, Exploratory Factor Analysis, and Internal consistency were conducted. Results provide the optimized version of the TMHQ-depression that contain 9 items. The 9 items are categorized into four factors which are suicidal ideation, sleep problems, anhedonic, and guilty feelings. Internal consistency analysis shows that this short version of the TMHQ-depression has good to excellent reliability (Cronbach's alpha >.80). The findings suggest that this optimized TMHQ-depression questionnaire holds a good psychometric property and can be used for depression detection.

Handwritten numeral recognition is a classical and important task in the computer vision area. We propose two novel deep learning models for this task, which combine the edge extraction method and Siamese/Triple network structures. We evaluate the models on seven handwritten numeral datasets and the results demonstrate both the simplicity and effectiveness of our models, comparing to baseline methods.

The advancement of multicore technology has made hundreds or even thousands of cores processor on a single chip possible. However, on a larger scale multicore, a hardware-based cache coherency mechanism becomes overwhelmingly complicated, hot, and expensive. Therefore, we propose a software coherence scheme managed by a parallelizing compiler for shared-memory multicore systems without a hardware cache coherence mechanism. Our proposed method is simple and efficient. It is built into OSCAR automatic parallelizing compiler. The OSCAR compiler parallelizes the coarse grain task, analyzes stale data and line sharing in the program, then solves those problems by simple program restructuring and data synchronization. Using our proposed method, we compiled 10 benchmark programs from SPEC2000, SPEC2006, NAS Parallel Benchmark (NPB), and MediaBench II. The compiled binaries then are run on Renesas RP2, an 8 cores SH-4A processor, and a custom 8-core Altera Nios II system on Altera Arria 10 FPGA. The cache coherence hardware on the RP2 processor is only available for up to 4 cores. The RP2's cache coherence hardware can also be turned off for non-coherence cache mode. The Nios II multicore system does not have any hardware cache coherence mechanism; therefore, running a parallel program is difficult without any compiler support. The proposed method performed as good as or better than the hardware cache coherence scheme while still provided the correct result as the hardware coherence mechanism. This method allows a massive array of shared memory CPU cores in an HPC setting or a simple non-coherent multicore embedded CPU to be easily programmed. For example, on the RP2 processor, the proposed software-controlled non-coherent-cache (NCC) method gave us 2.6 times speedup for SPEC 2000 “equake” with 4 cores against sequential execution while got only 2.5 times speedup for 4 cores MESI hardware coherent control. Also, the software coherence control gave us 4.4 times speedup for 8 cores with no hardware coherence mechanism available.

In this paper, we propose an effective and robust method of spatial feature extraction for acoustic scene analysis utilizing partially synchronized and/or closely located distributed microphones. In the proposed method, a new cepstrum feature utilizing a graph-based basis transformation to extract spatial information from distributed microphones, while taking into account whether any pairs of microphones are synchronized and/or closely located, is introduced. Specifically, in the proposed graph-based cepstrum, the log-amplitude of a multichannel observation is converted to a feature vector utilizing the inverse graph Fourier transform, which is a method of basis transformation of a signal on a graph. Results of experiments using real environmental sounds show that the proposed graph-based cepstrum robustly extracts spatial information with consideration of the microphone connections. Moreover, the results indicate that the proposed method more robustly classifies acoustic scenes than conventional spatial features when the observed sounds have a large synchronization mismatch between partially synchronized microphone groups.

The decentralized energy network is one of the promising solutions as a next-generation power grid. In this system, each house has a photovoltaic (PV) panel as a renewable energy source and a battery which is an essential component to balance between generation and demand. The common objective of the battery management on such systems is to minimize only the purchased energy from a power company, but battery degradation caused by charge/discharge cycles is also a serious problem. This paper proposes a State-of-Health (SOH) aware system-level battery management methodology for the decentralized energy network. The power distribution problem is often solved with mixed integer programming (MIP), and the proposed MIP formulation takes into account the SOH model. In order to minimize the purchased energy and reduce the battery degradation simultaneously, the optimization problem is divided into two stages: 1) the purchased energy minimization, and 2) the battery aging factor reducing, and the trade-off exploration between the purchased energy and the battery degradation is available. Experimental results show that the proposed method achieves the better trade-off and reduces the battery aging cost by 14% over the baseline method while keeping the purchased energy minimum.