• Toxicological Research
• /
• v.20 no.3
• /
• pp.213-223
• /
• 2004

This work aimed to study the effectiveness of cellular oxidative parameter (malondial-dehyde, protein carbonyl, and 8-hydroxy-2'deoxyguanosine). The experimental groups were aluminum treated rats and control rats. Aluminum treatd rats were given intraperitoneally aluminum nitrate nonahydrate ($Al^{3+}$, 0.2 mmol/kg) daily for 30 days except Sunday. Control rats were injected 1 ml of saline. After the dose, rats were decapitated and hippocampus and cerebral cortex were removed. The measured parameters were tissue malondialdehyde (MDA, index of lipid peroxidation), protein carbonyl (index of protein oxidation), 8-hydroxy-2'-deoxy-guanosine (8-OHdG, index of DNA oxidation), reduced glutathione (GSH) levels as well as glutathione reductase (GR) and catalase. AI concentrations in the tissues were also measured. All results were corrected by tissue protein levels. The results were as followed; 1. The concentrations of AI in the cortex and hippocampus were significantly higher in the AI-treated rats than in the control rats. 2. Antioxidative enzyme's activity, catalase and GR, were significantly higher in the AI-treated rats than the control rats. GSH levels were also higher in the AI-treated rats. 3. MDA, protein carbonyl, and 8-OHdG concentration of AI-treated rats were significantly higher than those of control rats. 4. The concentrations of antioxidants, and oxidative stress parameter were correlated with the concentrations of AI in hippocampus and cerebral cortex. Catalase and GR activity were also correlated with the concentration of AI. Based on these results, it can be suggested that intraperitoneally injected AI was accumulated in the brain and induced the increase of antioxidant levels and antioxidative enzyme activity. Also, the oxidative products of cellular macromolecules are significantly related to tissue AI concentration. Therefore MDA, protein carbonyl, and 8-OHdG are useful markers for oxidative stress on cellular macromolecules.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.11
• /
• pp.19-27
• /
• 2022

In this study, we propose a language and platform to describe and manage the MLOps(Machine Learning Operations) workflow for time series data anomaly detection. Time series data is collected in many fields, such as IoT sensors, system performance indicators, and user access. In addition, it is used in many applications such as system monitoring and anomaly detection. In order to perform prediction and anomaly detection of time series data, the MLOps platform that can quickly and flexibly apply the analyzed model to the production environment is required. Thus, we developed Python-based AI/ML Modeling Language (AMML) to easily configure and execute MLOps workflows. Python is widely used in data analysis. The proposed MLOps platform can extract and preprocess time series data from various data sources (R-DB, NoSql DB, Log File, etc.) using AMML and predict it through a deep learning model. To verify the applicability of AMML, the workflow for generating a transformer oil temperature prediction deep learning model was configured with AMML and it was confirmed that the training was performed normally.

• Journal of Information Science Theory and Practice
• /
• v.10 no.spc
• /
• pp.143-153
• /
• 2022

With the development of networks and the increase in the number of network devices, the number of cyber attacks targeting them is also increasing. Since these cyber-attacks aim to steal important information and destroy systems, it is necessary to minimize social and economic damage through early detection and rapid response. Many studies using machine learning (ML) and artificial intelligence (AI) have been conducted, among which payload learning is one of the most intuitive and effective methods to detect malicious behavior. In this study, we propose a preprocessing method to maximize the performance of the model when learning the payload in term units. The proposed method constructs a high-quality learning data set by eliminating unnecessary noise (stopwords) and preserving important features in consideration of the machine language and natural language characteristics of the packet payload. Our method consists of three steps: Preserving significant special characters, Generating a stopword list, and Class label refinement. By processing packets of various and complex structures based on these three processes, it is possible to make high-quality training data that can be helpful to build high-performance ML/AI models for security monitoring. We prove the effectiveness of the proposed method by comparing the performance of the AI model to which the proposed method is applied and not. Forthermore, by evaluating the performance of the AI model applied proposed method in the real-world Security Operating Center (SOC) environment with live network traffic, we demonstrate the applicability of the our method to the real environment.

• Journal of Embryo Transfer
• /
• v.32 no.3
• /
• pp.159-163
• /
• 2017

The artificial insemination (AI) is one of the best assisted reproductive technologies for increasing reproductive capacity and facilitating the genetic improvement in farm animals. AI has been used in Uganda for over 60 years, but a small population of the total herd has been used. This study was conducted to investigate the efficacy of AI with estrus synchronization technique and to propose ways of improving the productivity of dairy farms through AI services in Uganda. In total, 78 cows from 11 dairy farms were selected for timed-AI. Synchronization was performed according to the ovsynch programs followed by AI using frozen semen from Korean Holstein (0.5 ml straws). Pregnancy rate was varying among farms (0-50%) and the overall pregnancy rate was 28.2%. Cows in luteal phase at the time of treatment was 40.0% whereas that in follicular phase was 20.8%. After treatment, cows that showed normal estrus signal were 45.5% (25/55). Abnormal estrus was categorized into pre-estrus (9.1%), cystic ovaries (21.8%), anestrus (18.2%) and delayed ovulation (5.5%), respectively. These results imply that an assured protocol for timed-AI should be developed to improve the productivity of dairy farms through AI services in Uganda.

• Korean Journal of Veterinary Service
• /
• v.31 no.3
• /
• pp.397-414
• /
• 2008

In order to evaluate conception rate of Hanwoo in northwestern region of Gyeongsang-nam-do, we investigated conception rate and reduction of reproductive disorder rate after artificial insemination (AI) in 1,000 heads of breeding cows, This study showed that 80.9% of cows were classified as fertility after 1st and 2nd AI. For a accurate pregnancy diagnosis with practicing ovariectomy and histeotomy, we comparatively investigated each of 80 slaughtered cows, including 30 of non-pregnancy, and used enzyme-linked immunosorbent assay (ELISA) for estimation of plasma progesterone concentration and serum luteal hormone. The mean diameter of non-pregnant corpus luteum is $18.9{\pm}4.2{\times}15.6{\pm}3.6 mm$ and that of pregnant corpus luteum is $22.5{\pm}2.7{\times}18.7{\pm}2.9 mm$. This indicates that corpus luteum is more developed in the ovary of pregnant than non-pregnant cows (P<0.05). The diameter of pregnant corpus luteum according to the stage of pregnancy showed $21.3{\pm}2.4{\pm}18.4{\pm}2.6 mm$ in early stage (1-3 month), $23.4{\pm}2.8{\times}19.1{\pm}2.7 mm$ in middle stage (4-6 month) and $22.8{\pm}3.0{\times}18.8{\pm}2.4mm$, in last stage (7-9 month). This indicates that corpus luteum in middle and last stage is more significantly developed than that of early stage(P<0.05). The mean plasma progesterone concentration of cows showing size of non-pregnant corpus luteum was $4.58{\pm}0.92ng/ml$ and that of pregnant corpus luteum $8.26{\pm}0.98ng/ml$. Thus, it was more significantly increased in pregnant corpus luteum(P<0.02).. However, it was low to $0.58{\pm}0.39ng/ml$. in estrus (corpus albicans). The plasma progesterone concentration according to gestation period was high in proportion to the degree of development in corpus luteum and more significantly increased (P<0.05) and maintained in middle and last state than early state. The concentration was sharply decreased to $0.56{\pm}0.32ng/ml$ at parturition. As a consequence, we can practice the early pregnancy diagnosis by confirming non-pregnancy when the mean plasma progesterone concentration is below 1ng/ml 19 to 22 days after AI and this can be available to diagnose reproductive disorder.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.34 no.2
• /
• pp.113-119
• /
• 1989

In 1988 the tolerance of corn, sorghum, sorghum-sudangrass. and pearl millet hybrids to simazine and alachlor was tested in field during the growing season and pots during the summer and fall. In field and summer pot experiments(Exp.) the above mentioned four crops were tested at the ratios of simazine WP (50% ai, g/10a) ; alachlor EC (43.7% ai. ml/10a) of 130: 0, 100: 0, 70: 200, 0: 300 and 0: 400 and a sorghum hybrid was tested at 0, 50, 100, 200, 300, and 400ml/10a of alachlor and 70g/10a of simazine+ 200ml/10a of alachlor in fall pot Exp. In corn emergence rate, percent stand, plant height of seedlings, and dry matter(DM) yield were not affected by simazine and alachlor in all Exps. In sorghum and sorghum-sudangrass early growth and DM yield were not affected by simazine and alachlor in field Exp. In contrast, simazine reduced height and dry weight of seedlings slightly without any deterimental effects on emergence and survival rates. but alachlor reduced survival rate, plant height, and dry weight of seedlings significantly in summer pot Exp. In fall Exp. alachlor did not affect emergence rate of a sorghum hybrid, but survival rate, plant height, and dry weight of seedlings reduced with increased levels of alachlor when applied higher than 100ml/10a. In pearl millet simazine did not affect emergence rate, plant height, and DM yield in field, but reduced survival rate, plant height, and dry weight of seedlings in summer pot Exp. However, alachlor reduced DM yield significantly due to a lower percent stand even in the field. In summer pot Exp. although emeregence rate was slightly reduced, all seedlings were dead after emergence. Simazine did not control grasses such as Digitaria sanguinalis, Setaria viridis, Echinochloa crusgalli effectively, but controlled broadleaf weeds. Alachlor controlled all grasses, Porluraca oleracea, and Amaranthus mangostanus, but did not control Acalypha australis and Chenopodium album. A combination of simazine and alachlor controlled weeds more effectively than either simazine or alachlor alone.

• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.672-678
• /
• 2021

New malware continues to increase and become advanced by every year. Although various studies are going on executable files to diagnose malicious codes, it is difficult to detect attacks that internalize malicious code threats in emails by exploiting non-executable document files, malicious URLs, and malicious macros and JS in documents. In this paper, we introduce a method of analyzing malicious code for email security through proactive detection and blocking of malicious email attacks, and propose a method for determining whether a non-executable document file is malicious based on AI. Among various algorithms, an efficient machine learning modeling is choosed, and an ML workflow system to diagnose malicious code using Kubeflow is proposed.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.924-926
• /
• 2021

AI (Artificial Intelligence) is being utilized in various fields and services to give convenience to human life. Unfortunately, there are many security vulnerabilities in today's ML (Machine Learning) systems, causing various privacy concerns as some AI models need individuals' private data to train them. Such concerns lead to the interest in ML systems which can preserve the privacy of individuals' data. This paper introduces the latest research on various attacks that infringe data privacy and the corresponding defense techniques.

• Electronics and Telecommunications Trends
• /
• v.38 no.6
• /
• pp.95-106
• /
• 2023

As artificial intelligence has become commonplace in various fields, the transparency of AI in its development and implementation has become an important issue. In safety-critical areas, the eXplainable and/or understandable of artificial intelligence is being actively studied. On the other hand, machine learning have been applied to the intelligence of self-organizing network (SON), but transparency in this application has been neglected, despite the critical decision-makings in the operation of mobile communication systems. We describes concepts of eXplainable machine learning (ML), along with research trends, major issues, and research directions. After summarizing the ML research on SON, research directions are analyzed for explainable ML required in intelligent SON of beyond 5G and 6G communication.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.2
• /
• pp.205-211
• /
• 2024

The purpose of this study is to implement an agent that intelligently performs tracking and movement through reinforcement learning using the Unity and ML-Agents. In this study, we conducted an experiment to compare the learning performance between training one agent in a single learning simulation environment and parallel training of several agents simultaneously in a multi-learning simulation environment. From the experimental results, we could be confirmed that the parallel training method is about 4.9 times faster than the single training method in terms of learning speed, and more stable and effective learning occurs in terms of learning stability.