• Journal of Internet Computing and Services
• /
• v.22 no.3
• /
• pp.9-16
• /
• 2021

This study aims to develop a human activity recognition (HAR) system as a Deep-Learning (DL) classification model, distinguishing various human activities. We solely rely on the signals from a wristband accelerometer worn by a person for the user's convenience. 3-axis sequential acceleration signal data are gathered within a predefined time-window-slice, and they are used as input to the classification system. We are particularly interested in developing a Deep-Learning model that can outperform conventional machine learning classification performance. A total of 13 activities based on the laboratory experiments' data are used for the initial performance comparison. We have improved classification performance using the Convolutional Neural Network (CNN) combined with an auto-encoder feature reduction and parameter tuning. With various publically available HAR datasets, we could also achieve significant improvement in HAR classification. Our CNN model is also compared against Recurrent-Neural-Network(RNN) with Long Short-Term Memory(LSTM) to demonstrate its superiority. Noticeably, our model could distinguish both general activities and near-identical activities such as sitting down on the chair and floor, with almost perfect classification accuracy.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.7
• /
• pp.1088-1097
• /
• 2021

Vibration data of mechanical equipment inevitably have noise. This noise adversely af ects the maintenance of mechanical equipment. Accordingly, the performance of a learning model depends on how effectively the noise of the data is removed. In this study, the noise of the data was removed using the Denoising Auto Encoder (DAE) technique which does not include the characteristic extraction process in preprocessing time series data. In addition, the performance was compared with that of the Wavelet Transform, which is widely used for machine signal processing. The performance comparison was conducted by calculating the failure detection rate. For a more accurate comparison, a classification performance evaluation criterion, the F-1 Score, was calculated. Failure data were detected using the One-Class SVM technique. The performance comparison, revealed that the DAE technique performed better than the Wavelet Transform technique in terms of failure diagnosis and error rate.

• Smart Media Journal
• /
• v.13 no.6
• /
• pp.9-15
• /
• 2024

In recent trends, there has been an increase in 'Qshing' attacks, a hybrid form of phishing that exploits fake QR (Quick Response) codes impersonating government agencies to steal personal and financial information. Particularly, this attack method is characterized by its stealthiness, as victims can be redirected to phishing pages or led to download malicious software simply by scanning a QR code, making it difficult for them to realize they have been targeted. In this paper, we have developed a classification technique utilizing machine learning algorithms to identify the maliciousness of URLs embedded in QR codes, and we have explored ways to integrate this with existing QR code readers. To this end, we constructed a dataset from 128,587 malicious URLs and 428,102 benign URLs, extracting 35 different features such as protocol and parameters, and used AutoML to identify the optimal algorithm and hyperparameters, achieving an accuracy of approximately 87.37%. Following this, we designed the integration of the trained classification model with existing QR code readers to implement a service capable of countering Qshing attacks. In conclusion, our findings confirm that deriving an optimized algorithm for classifying malicious URLs in QR codes and integrating it with existing QR code readers presents a viable solution to combat Qshing attacks.

• Proceedings of the KSRS Conference
• /
• 2006.03a
• /
• pp.117-120
• /
• 2006

Heavy rainfall events are occurred exceedingly various forms by a complex interaction between synoptic, dynamic and atmospheric stability. As the results, quantitative precipitation forecast is extraordinary difficult because it happens locally in a short time and has a strong spatial and temporal variations. GOES-9 imagery data provides continuous observations of the clouds in time and space at the right resolution. In this study, an power-law type algorithm(KAE: Korea auto estimator) for estimating rainfall based on the rainfall type was developed using geostationary meteorological satellite data. GOES-9 imagery and automatic weather station(AWS) measurements data were used for the classification of rainfall types and the development of estimation algorithm. Subjective and objective classification of rainfall types using GOES-9 imagery data and AWS measurements data showed that most of heavy rainfalls are occurred by the convective and mired type. Statistical analysis between AWS rainfall and GOES-IR data according to the rainfall types showed that estimation of rainfall amount using satellite data could be possible only for the convective and mixed type rainfall. The quality of KAE in estimating the rainfall amount and rainfall area is similar or slightly superior to the National Environmental Satellite Data and Information Service's auto-estimator(NESDIS AE), especially for the multi cell convective and mixed type heavy rainfalls. Also the high estimated level is denoted on the mature stage as well as decaying stages of rainfall system.

• The Bulletin of The Korean Astronomical Society
• /
• v.39 no.1
• /
• pp.75.2-75.2
• /
• 2014

We have developed a set of daily solar flare peak flux forecast models using the multiple linear regression (MLR), the auto regression (AR), and artificial neural network (ANN) methods. We consider input parameters as solar activity data from January 1996 to December 2013 such as sunspot area, X-ray flare peak flux, weighted total flux $T_F=1{\times}F_C+10{\times}F_M+100{\times}F_X$ of previous day, mean flare rates of a given McIntosh sunspot group (Zpc), and a Mount Wilson magnetic classification. We compute the hitting rate that is defined as the fraction of the events whose absolute differences between the observed and predicted flare fluxes in a logarithm scale are ${\leq}$ 0.5. The best three parameters related to the observed flare peak flux are as follows: weighted total flare flux of previous day (r=0.5), Mount Wilson magnetic classification (r=0.33), and McIntosh sunspot group (r=0.3). The hitting rates of flares stronger than the M5 class, which is regarded to be significant for space weather forecast, are as follows: 30% for the auto regression method and 69% for the neural network method.

• Journal of Auto-vehicle Safety Association
• /
• v.12 no.4
• /
• pp.61-69
• /
• 2020

As a sales volume of autonomous vehicle continually grows up, regulations on this new technology are being introduced around the world. For example, safety standards for the Level 3 automated driving system was promulgated in December 2019 by the Ministry of Land, Infrastructure and Transport of Korean government. In order to promote the development of autonomous vehicle technology and ensure its safety simultaneously, the regulations on the automated driving systems should be phased in to keep pace with technology progress and market expansion. However, according to SAE J3016, which is well known to classify the level of the autonomous vehicle technologies, the description for classification is rather abstract. Therefore it is necessary to describe the automated driving system in more detail in terms of the 'Level.' In this study, the functions and characteristics of automated driving system are carefully classified at each level based on the commentary in the Informal Working Group (IWG) of the UN WP29. In particular, regarding the Level 4, technical issues are characterized with respect to vehicle tasks, driver tasks, system performance and regulations. The important features of the autonomous vehicles to meet Level 4 are explored on the viewpoints of driver replacement, emergency response and connected driving performance.

• Journal of Auto-vehicle Safety Association
• /
• v.16 no.1
• /
• pp.35-41
• /
• 2024

In order to verify autonomous driving scenarios and safety, a lot of driving and accident data is needed, so various organizations are conducting classification and analysis of traffic accident types. In this study, it was determined that accident recording devices such as EDR (Event Data Recorder) and DSSAD (Data Storage System for Automated Driving) would become an objective standard for analyzing the causes of autonomous vehicle accidents, and traffic accidents that occurred from 2015 to 2020 were analyzed. Using the database system of IGLAD (Initiative for the Global Harmonization of Accident Data), approximately 360 accident data of EDR-equipped vehicles were classified and their characteristics were analyzed by comparing them with accident types of ADAS (Advanced Driver Assistance System)-equipped vehicles. It will be used to develop autonomous vehicle accident investigation guidelines in the future.

• International Journal of Computer Science & Network Security
• /
• v.23 no.8
• /
• pp.49-62
• /
• 2023

There has been a rapid increase in the creation and alteration of new malware samples which is a huge financial risk for many organizations. There is a huge demand for improvement in classification and detection mechanisms available today, as some of the old strategies like classification using mac learning algorithms were proved to be useful but cannot perform well in the scalable auto feature extraction scenario. To overcome this there must be a mechanism to automatically analyze malware based on the automatic feature extraction process. For this purpose, the dynamic analysis of real malware executable files has been done to extract useful features like API call sequence and opcode sequence. The use of different hashing techniques has been analyzed to further generate images and convert them into image representable form which will allow us to use more advanced classification approaches to classify huge amounts of images using deep learning approaches. The use of deep learning algorithms like convolutional neural networks enables the classification of malware by converting it into images. These images when fed into the CNN after being converted into the grayscale image will perform comparatively well in case of dynamic changes in malware code as image samples will be changed by few pixels when classified based on a greyscale image. In this work, we used VGG-16 architecture of CNN for experimentation.

• Journal of Korea Game Society
• /
• v.16 no.6
• /
• pp.163-172
• /
• 2016

This study analyzes the attitudes of users who use the automatic-play function of mobile-games. It also observes changes in different emotions to enjoyment. For the analysis, the Q methodology appropriate for the subjectivity study was used, and it classified two types of attitudes based on the analysis. After the classification, it defined the characteristic of each type based on Lazzaro's theory on classification of fun. Types with high rate of convenience feature dependency utilize automatic-play feature to play games for collecting and growth, and they prefer managing the game outdoors. Another type enjoyed the experience of problem solving through games. They want the process to be a good skill to possess in game-play. There was also a tendency to perceive oneself as the game's character. Both types were responded positively to using automatic-play to play mobile-games. This is expected to be used as important data for game production, leading to usability evaluation for analyzing the emotional sensitivity of mobile game users.

• Journal of Information Processing Systems
• /
• v.16 no.1
• /
• pp.6-29
• /
• 2020

Biometrics identification using multiple modalities has attracted the attention of many researchers as it produces more robust and trustworthy results than single modality biometrics. In this paper, we present a novel multimodal recognition system that trains a deep learning network to automatically learn features after extracting multiple biometric modalities from a single data source, i.e., facial video clips. Utilizing different modalities, i.e., left ear, left profile face, frontal face, right profile face, and right ear, present in the facial video clips, we train supervised denoising auto-encoders to automatically extract robust and non-redundant features. The automatically learned features are then used to train modality specific sparse classifiers to perform the multimodal recognition. Moreover, the proposed technique has proven robust when some of the above modalities were missing during the testing. The proposed system has three main components that are responsible for detection, which consists of modality specific detectors to automatically detect images of different modalities present in facial video clips; feature selection, which uses supervised denoising sparse auto-encoders network to capture discriminative representations that are robust to the illumination and pose variations; and classification, which consists of a set of modality specific sparse representation classifiers for unimodal recognition, followed by score level fusion of the recognition results of the available modalities. Experiments conducted on the constrained facial video dataset (WVU) and the unconstrained facial video dataset (HONDA/UCSD), resulted in a 99.17% and 97.14% Rank-1 recognition rates, respectively. The multimodal recognition accuracy demonstrates the superiority and robustness of the proposed approach irrespective of the illumination, non-planar movement, and pose variations present in the video clips even in the situation of missing modalities.