• Journal of Convergence for Information Technology
• /
• v.12 no.2
• /
• pp.47-53
• /
• 2022

Recently, with the change of the intelligent security paradigm, study to apply various information generated from various information security systems to AI-based anomaly detection is increasing. Therefore, in this study, in order to convert log-like time series data into a vector, which is a numerical feature, the CBOW and Skip-gram inference methods of deep learning-based Word2Vec model and statistical method based on the coincidence frequency were used to transform the published ADFA system call data. In relation to this, an experiment was carried out through conversion into various embedding vectors considering the dimension of vector, the length of sequence, and the window size. In addition, the performance of the embedding methods used as well as the detection performance were compared and evaluated through GRU-based anomaly detection model using vectors generated by the embedding model as an input. Compared to the statistical model, it was confirmed that the Skip-gram maintains more stable performance without biasing a specific window size or sequence length, and is more effective in making each event of sequence data into an embedding vector.

• Smart Media Journal
• /
• v.12 no.9
• /
• pp.89-94
• /
• 2023

Recommender systems provide users with the most favorable items by analyzing explicit or implicit feedback of users on items. Recently, as the size of deep-learning-based models employed in recommender systems has increased, many studies have focused on reducing inference time while maintaining high recommendation accuracy. As one of them, a study on recommender systems with a knowledge distillation (KD) technique is actively conducted. By KD, a small-sized model (i.e., student) is trained through knowledge extracted from a large-sized model (i.e., teacher), and then the trained student is used as a recommendation model. Existing studies on KD for recommender systems have been mainly performed only for implicit feedback settings. Thus, in this paper, we try to investigate the performance and accuracy when applied to explicit feedback settings. To this end, we leveraged a total of five state-of-the-art KD methods and three real-world datasets for recommender systems.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.3
• /
• pp.151-158
• /
• 2024

Although on-device artificial intelligence (AI) has gained attention to diagnosing machine faults in real time, most previous studies did not consider the model retraining and redeployment processes that must be performed in real-world industrial environments. Our study addresses this challenge by proposing an on-device AI-based real-time machine fault diagnosis system that utilizes continual learning. Our proposed system includes a lightweight convolutional neural network (CNN) model, a continual learning algorithm, and a real-time monitoring service. First, we developed a lightweight 1D CNN model to reduce the cost of model deployment and enable real-time inference on the target edge device with limited computing resources. We then compared the performance of five continual learning algorithms with three public bearing fault datasets and selected the most effective algorithm for our system. Finally, we implemented a real-time monitoring service using an open-source data visualization framework. In the performance comparison results between continual learning algorithms, we found that the replay-based algorithms outperformed the regularization-based algorithms, and the experience replay (ER) algorithm had the best diagnostic accuracy. We further tuned the number and length of data samples used for a memory buffer of the ER algorithm to maximize its performance. We confirmed that the performance of the ER algorithm becomes higher when a longer data length is used. Consequently, the proposed system showed an accuracy of 98.7%, while only 16.5% of the previous data was stored in memory buffer. Our lightweight CNN model was also able to diagnose a fault type of one data sample within 3.76 ms on the Raspberry Pi 4B device.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.4
• /
• pp.186-191
• /
• 2023

Data processing techniques play a critical role in decision-making, including handling missing and outlier data, prediction, and recommendation models. This requires a clear explanation of the validity, reliability, and accuracy of all processes and results. In addition, it is necessary to solve data problems through explainable models using decision trees, inference, etc., and proceed with model lightweight by considering various types of learning. The multi-layer mining classification method that applies the sixth principle is a method that discovers multidimensional relationships between variables and attributes that occur frequently in transactions after data preprocessing. This explains how to discover significant relationships using mining on transactions and model the data through regression analysis. It develops scalable models and logistic regression models and proposes mining techniques to generate class labels through data cleansing, relevance analysis, data transformation, and data augmentation to make explanatory decisions.

• Korean Journal of Culture and Social Issue
• /
• v.26 no.1
• /
• pp.25-45
• /
• 2020

Intentionality judgement in criminal cases is a core area of fact finding that is root of guilty and sentencing judgment on the defendant. However, the third party is not sure the intentionality because it reflects subjective aspect of agent. Thus, mechanism behind intentionality judgment is an important factor to be properly understood by the academia and the criminal justice system. However, previous studies regarding intentionality judgment models have shown inconsistent results. Mental-state models proposed foreseeability(belief) and desire of agent at the time of the offence as key factors in intentionality judgment. These factors consistent with central things on intentionality judgment in criminal law. However, key factors in moral-evaluation models are blameworthiness of agent and badness of outcome reflected on the consequent aspect of act. Recently, deep-self concordance model emerged suggesting important factors on intentionality judgment are not mental states and moral evaluations but individual's deep-self. However, these models are limited in that they do not consider the important features of criminal cases, that the consequence of the case is inevitably negative, and therefore the actor who is a party to legal punishment rarely expresses his or her mental state at the time of the act. Therefore, this study suggests that, based on the existing intentionality judgment studies and the characteristics of the criminal case, the inference about who the agent was originally will play a key role in judging the intentionality in the criminal case. This is the moral-character model. Futhermore, In this regard, this study discussed what the media and criminal justice institutions should keep in mind and the directions for future research.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.18 no.1
• /
• pp.135-141
• /
• 2018

In this paper, we describe design and implementation of optimal smart home control system. Recent developments in technologies such as sensors and communication have enabled the Internet of Things to control a wide range of objects, such as light bulbs, socket-outlet, or clothing. Many businesses rely on the launch of collaborative services between them. However, traditional IoT systems often support a single protocol, although data is transmitted across multiple protocols for end-to-end devices. In addition, depending on the manufacturer of the Internet of things, there is a dedicated application and it has a high degree of complexity in registering and controlling different IoT devices for the internet of things. ARIoT system, special marking points and edge extraction techniques are used to detect objects, but there are relatively low deviations depending on the sampling data. The proposed system implements an IoT gateway of object based on OneM2M to compensate for existing problems. It supports diverse protocols of end to end devices and supported them with a single application. In addition, devices were learned by using deep learning in the artificial intelligence field and improved object recognition of existing systems by inference and detection, reducing the deviation of recognition rates.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.4
• /
• pp.29-35
• /
• 2020

In this paper, we propose a new structured entity recognition DeNERT model. Recently, the field of natural language processing has been actively researched using pre-trained language representation models with a large amount of corpus. In particular, the named entity recognition, which is one of the fields of natural language processing, uses a supervised learning method, which requires a large amount of training dataset and computation. Reinforcement learning is a method that learns through trial and error experience without initial data and is closer to the process of human learning than other machine learning methodologies and is not much applied to the field of natural language processing yet. It is often used in simulation environments such as Atari games and AlphaGo. BERT is a general-purpose language model developed by Google that is pre-trained on large corpus and computational quantities. Recently, it is a language model that shows high performance in the field of natural language processing research and shows high accuracy in many downstream tasks of natural language processing. In this paper, we propose a new named entity recognition DeNERT model using two deep learning models, DQN and BERT. The proposed model is trained by creating a learning environment of reinforcement learning model based on language expression which is the advantage of the general language model. The DeNERT model trained in this way is a faster inference time and higher performance model with a small amount of training dataset. Also, we validate the performance of our model's named entity recognition performance through experiments.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.12 no.5
• /
• pp.467-474
• /
• 2019

In this paper, we propose a novel method to improve FPS while maintaining the accuracy of YOLO v2 model in NVIDIA Jetson TX1. In general, in order to reduce the amount of computation, a conversion to an integer operation or reducing the depth of a network have been used. However, the accuracy of recognition can be deteriorated. So, we use methods to reduce computation and memory consumption through adjustment of the filter size and integrated computation of the network The first method is to replace the $3{\times}3$ filter with a $1{\times}1$ filter, which reduces the number of parameters to one-ninth. The second method is to reduce the amount of computation through CBR (Convolution-Add Bias-Relu) among the inference acceleration functions of TensorRT, and the last method is to reduce memory consumption by integrating repeated layers using TensorRT. For the simulation results, although the accuracy is decreased by 1% compared to the existing YOLO v2 model, the FPS has been improved from the existing 3.9 FPS to 11 FPS.

• Journal of Advanced Navigation Technology
• /
• v.26 no.4
• /
• pp.211-218
• /
• 2022

Continuous wave (CW) radar has the advantage of reliability and accuracy compared to other sensors such as camera and lidar. In addition, binarized neural network (BNN) has a characteristic that dramatically reduces memory usage and complexity compared to other deep learning networks. Therefore, this paper proposes binarized neural network based human identification and motion classification system using CW radar. After receiving a signal from CW radar, a spectrogram is generated through a short-time Fourier transform (STFT). Based on this spectrogram, we propose an algorithm that detects whether a person approaches a radar. Also, we designed an optimized BNN model that can support the accuracy of 90.0% for human identification and 98.3% for motion classification. In order to accelerate BNN operation, we designed BNN hardware accelerator on field programmable gate array (FPGA). The accelerator was implemented with 1,030 logics, 836 registers, and 334.904 Kbit block memory, and it was confirmed that the real-time operation was possible with a total calculation time of 6 ms from inference to transferring result.

• The Journal of Bigdata
• /
• v.8 no.1
• /
• pp.99-109
• /
• 2023

Recently, not only in Korea but also around the world, we have been experiencing constant disasters such as typhoons, wildfires, and heavy rains. The property damage caused by typhoons and heavy rain in South Korea alone has exceeded 1 trillion won. These disasters have resulted in significant loss of life and property damage, and the recovery process will also take a considerable amount of time. In addition, the government's contingency funds are insufficient for the current situation. To prevent and effectively respond to these issues, it is necessary to collect and analyze accurate data in real-time. However, delays and data loss can occur depending on the environment where the sensors are located, the status of the communication network, and the receiving servers. In this paper, we propose a two-stage hybrid situation analysis and prediction algorithm that can accurately analyze even in such communication network conditions. In the first step, data on river and stream levels are collected, filtered, and refined from diverse sensors of different types and stored in a bigdata. An AI rule-based inference algorithm is applied to analyze the crisis alert levels. If the rainfall exceeds a certain threshold, but it remains below the desired level of interest, the second step of deep learning image analysis is performed to determine the final crisis alert level.