• 제목/요약/키워드: Machine Learning and Artificial Intelligence

검색결과 747건 처리시간 0.025초

인공지능 프로세서 기술 동향 (AI Processor Technology Trends)

  • 권영수
    • 전자통신동향분석
    • /
    • 제33권5호
    • /
    • pp.121-134
    • /
    • 2018
  • The Von Neumann based architecture of the modern computer has dominated the computing industry for the past 50 years, sparking the digital revolution and propelling us into today's information age. Recent research focus and market trends have shown significant effort toward the advancement and application of artificial intelligence technologies. Although artificial intelligence has been studied for decades since the Turing machine was first introduced, the field has recently emerged into the spotlight thanks to remarkable milestones such as AlexNet-CNN and Alpha-Go, whose neural-network based deep learning methods have achieved a ground-breaking performance superior to existing recognition, classification, and decision algorithms. Unprecedented results in a wide variety of applications (drones, autonomous driving, robots, stock markets, computer vision, voice, and so on) have signaled the beginning of a golden age for artificial intelligence after 40 years of relative dormancy. Algorithmic research continues to progress at a breath-taking pace as evidenced by the rate of new neural networks being announced. However, traditional Von Neumann based architectures have proven to be inadequate in terms of computation power, and inherently inefficient in their processing of vastly parallel computations, which is a characteristic of deep neural networks. Consequently, global conglomerates such as Intel, Huawei, and Google, as well as large domestic corporations and fabless companies are developing dedicated semiconductor chips customized for artificial intelligence computations. The AI Processor Research Laboratory at ETRI is focusing on the research and development of super low-power AI processor chips. In this article, we present the current trends in computation platform, parallel processing, AI processor, and super-threaded AI processor research being conducted at ETRI.

A Study on Prediction of Baseball Game Based on Linear Regression

  • LEE, Kwang-Keun;HWANG, Seung-Ho
    • 한국인공지능학회지
    • /
    • 제7권2호
    • /
    • pp.13-17
    • /
    • 2019
  • Currently, the sports market continues to grow every year, and among them, professional baseball's entry income is larger than the rest of the professional league. In sports, strategies are used differently in different situations, and the analysis is based on data to decide which direction to implement. There is a part that a person misses in an analysis, and there is a possibility of a false analysis by subjective judgment. So, if this data analysis is done through artificial intelligence, the objective analysis is possible, and the strategy can be more rationalized, which helps to win the game. The most popular baseball to be applied to artificial intelligence to analyze athletes' strengths and weaknesses and then efficiently establish strategies to ease the competition. The data applied to the experiment were provided on the KBO official website, and the algorithms for forecasting applied linear regression. The results showed that the accuracy was 87%, and the standard error was ±5. Although the results of the experiment were not enough data, it would be possible to effectively use baseball strategies and predict the results of the game if the amount of data and regular data can be applied in the future.

오픈신경망 포맷을 이용한 기계학습 모델 변환 및 추론 (Model Transformation and Inference of Machine Learning using Open Neural Network Format)

  • 김선민;한병현;허준영
    • 한국인터넷방송통신학회논문지
    • /
    • 제21권3호
    • /
    • pp.107-114
    • /
    • 2021
  • 최근 다양한 분야에 인공지능 기술이 도입되고, 학계 관심이 늘어남에 따라 다양한 기계학습 모델들이 여러 프레임워크에서 운용되고 있다. 하지만 이러한 프레임워크들은 서로 다른 데이터 포맷을 가지고 있어, 상호운용성이 부족하며 이를 극복하기 위해 오픈 신경망 교환 포맷인 ONNX가 제안되었다. 본 논문에서는 여러 기계학습 모델을 ONNX로 변환하는 방법을 설명하고, 통합된 ONNX 포맷에서 기계학습 기법을 판별할 수 있는 알고리즘 및 추론 시스템을 제안한다. 또한, ONNX 변환 전·후 모델의 추론 성능을 비교하여 ONNX 변환 간 학습 결과의 손실이나 성능 저하가 없음을 보인다.

Implementation of Fund Recommendation System Using Machine Learning

  • Park, Chae-eun;Lee, Dong-seok;Nam, Sung-hyun;Kwon, Soon-kak
    • Journal of Multimedia Information System
    • /
    • 제8권3호
    • /
    • pp.183-190
    • /
    • 2021
  • In this paper, we implement a system for a fund recommendation based on the investment propensity and for a future fund price prediction. The investment propensity is classified by scoring user responses to series of questions. The proposed system recommends the funds with a suitable risk rating to the investment propensity of the user. The future fund prices are predicted by Prophet model which is one of the machine learning methods for time series data prediction. Prophet model predicts future fund prices by learning the parameters related to trend changes. The prediction by Prophet model is simple and fast because the temporal dependency for predicting the time-series data can be removed. We implement web pages for the fund recommendation and for the future fund price prediction.

CORRECT? CORECT!: Classification of ESG Ratings with Earnings Call Transcript

  • Haein Lee;Hae Sun Jung;Heungju Park;Jang Hyun Kim
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제18권4호
    • /
    • pp.1090-1100
    • /
    • 2024
  • While the incorporating ESG indicator is recognized as crucial for sustainability and increased firm value, inconsistent disclosure of ESG data and vague assessment standards have been key challenges. To address these issues, this study proposes an ambiguous text-based automated ESG rating strategy. Earnings Call Transcript data were classified as E, S, or G using the Refinitiv-Sustainable Leadership Monitor's over 450 metrics. The study employed advanced natural language processing techniques such as BERT, RoBERTa, ALBERT, FinBERT, and ELECTRA models to precisely classify ESG documents. In addition, the authors computed the average predicted probabilities for each label, providing a means to identify the relative significance of different ESG factors. The results of experiments demonstrated the capability of the proposed methodology in enhancing ESG assessment criteria established by various rating agencies and highlighted that companies primarily focus on governance factors. In other words, companies were making efforts to strengthen their governance framework. In conclusion, this framework enables sustainable and responsible business by providing insight into the ESG information contained in Earnings Call Transcript data.

생존분석에서의 기계학습 (Machine learning in survival analysis)

  • 백재욱
    • 산업진흥연구
    • /
    • 제7권1호
    • /
    • pp.1-8
    • /
    • 2022
  • 본 논문은 중도중단 데이터가 포함된 생존데이터의 경우 적용할 수 있는 기계학습 방법에 대해 살펴보았다. 우선 탐색적인 자료분석으로 각 특성에 대한 분포, 여러 특성들 간의 관계 및 중요도 순위를 파악할 수 있었다. 다음으로 독립변수에 해당하는 여러 특성들과 종속변수에 해당하는 특성(사망여부) 간의 관계를 분류문제로 보고 logistic regression, K nearest neighbor 등의 기계학습 방법들을 적용해본 결과 적은 수의 데이터이지만 통상적인 기계학습 결과에서와 같이 logistic regression보다는 random forest가 성능이 더 좋게 나왔다. 하지만 근래에 성능이 좋다고 하는 artificial neural network나 gradient boost와 같은 기계학습 방법은 성능이 월등히 좋게 나오지 않았는데, 그 이유는 주어진 데이터가 빅데이터가 아니기 때문인 것으로 판명된다. 마지막으로 Kaplan-Meier나 Cox의 비례위험모델과 같은 통상적인 생존분석 방법을 적용하여 어떤 독립변수가 종속변수 (ti, δi)에 결정적인 영향을 미치는지 살펴볼 수 있었으며, 기계학습 방법에 속하는 random forest를 중도중단 데이터가 포함된 생존데이터에도 적용하여 성능을 평가할 수 있었다.

Using Machine Learning Techniques for Accurate Attack Detection in Intrusion Detection Systems using Cyber Threat Intelligence Feeds

  • Ehtsham Irshad;Abdul Basit Siddiqui
    • International Journal of Computer Science & Network Security
    • /
    • 제24권4호
    • /
    • pp.179-191
    • /
    • 2024
  • With the advancement of modern technology, cyber-attacks are always rising. Specialized defense systems are needed to protect organizations against these threats. Malicious behavior in the network is discovered using security tools like intrusion detection systems (IDS), firewall, antimalware systems, security information and event management (SIEM). It aids in defending businesses from attacks. Delivering advance threat feeds for precise attack detection in intrusion detection systems is the role of cyber-threat intelligence (CTI) in the study is being presented. In this proposed work CTI feeds are utilized in the detection of assaults accurately in intrusion detection system. The ultimate objective is to identify the attacker behind the attack. Several data sets had been analyzed for attack detection. With the proposed study the ability to identify network attacks has improved by using machine learning algorithms. The proposed model provides 98% accuracy, 97% precision, and 96% recall respectively.

머신러닝기반 간 경화증 진단을 위한 웹 서비스 개발 (Development of Web Service for Liver Cirrhosis Diagnosis Based on Machine Learning)

  • 노시형;김지언;이충섭;김태훈;김경원;윤권하;정창원
    • 정보처리학회논문지:컴퓨터 및 통신 시스템
    • /
    • 제10권10호
    • /
    • pp.285-290
    • /
    • 2021
  • 의료분야에서 인공지능 기술을 도입한 질환 진단 및 예측 연구들이 활발하게 진행되고 있다. 의료영상기반의 인공지능 기술 적용에 가장 많이 활용되고 있는 질환 진단 및 예측에 대한 다양한 제품으로 출시되고 있다. 인공지능은 질병에 대한 진단, 양성과 악성으로 구분되는 질환의 구분, 질병의 위험도에 따른 구별이나 판독에 이용하기 위해 질환부위를 분리하는 등에 적용되고 있다. 최근에는 클라우드기술과 연계하여 서비스 제품으로 활용성이 높아지고 있다. 본 논문에서 다루는 질환 중에 간 질환은 통증이 적어 조기진단이 어려워 그 위험도가 매우 높은 질환이다. 이러한 질환 진단에 비침습적인 진단방법으로 의료영상기반으로 인공지능 기술을 도입하였다. 우리는 임상에서 가장 의미 있는 간 경화증 환자의 판독을 돕기 위한 웹 서비스 개발 내용을 기술한다. 그리고 웹서비스 프로세스를 보이고 각 프로세스의 구동 화면과 최종 결과화면을 보인다. 제안한 서비스를 통해 간 경화증을 조기에 진단하고, 빠른 치료를 통해 환자의 회복에 도움을 줄 수 있을 것으로 기대한다.

Dropout Genetic Algorithm Analysis for Deep Learning Generalization Error Minimization

  • Park, Jae-Gyun;Choi, Eun-Soo;Kang, Min-Soo;Jung, Yong-Gyu
    • International Journal of Advanced Culture Technology
    • /
    • 제5권2호
    • /
    • pp.74-81
    • /
    • 2017
  • Recently, there are many companies that use systems based on artificial intelligence. The accuracy of artificial intelligence depends on the amount of learning data and the appropriate algorithm. However, it is not easy to obtain learning data with a large number of entity. Less data set have large generalization errors due to overfitting. In order to minimize this generalization error, this study proposed DGA(Dropout Genetic Algorithm) which can expect relatively high accuracy even though data with a less data set is applied to machine learning based genetic algorithm to deep learning based dropout. The idea of this paper is to determine the active state of the nodes. Using Gradient about loss function, A new fitness function is defined. Proposed Algorithm DGA is supplementing stochastic inconsistency about Dropout. Also DGA solved problem by the complexity of the fitness function and expression range of the model about Genetic Algorithm As a result of experiments using MNIST data proposed algorithm accuracy is 75.3%. Using only Dropout algorithm accuracy is 41.4%. It is shown that DGA is better than using only dropout.

인공지능(Artificial Intelligence)과 대학수학교육 (Artificial Intelligence and College Mathematics Education)

  • 이상구;이재화;함윤미
    • 한국수학교육학회지시리즈E:수학교육논문집
    • /
    • 제34권1호
    • /
    • pp.1-15
    • /
    • 2020
  • 첨단 정보통신기술(ICT)인 인공지능(AI), 사물인터넷(IoT), 빅데이터(Big Data) 등이 사회와 경제 전반에 융합돼 혁신적인 변화가 일어나는 요즘, 헬스케어, 지능형 로봇, 가정용 인공지능 시스템(스마트홈), 공유자동차 등은 이미 우리 생활에 깊이 영향을 미치고 있다. 이미 오래전부터 공장에서는 로봇이 사람 대신 일을 하고 있으며(FA, OA), 인공지능 의사도 병원에서 활동을 하고 있고(Dr. Watson), 인공지능 스피커(기가지니)와 인공지능 비서인 구글 어시스턴트가 자연어생성을 하며 우리를 돕고 있다. 이제 인공지능을 이해하는 것은 필수가 되었으며, 인공지능을 이해하기 위해서 수학의 지식은 선택이 아니라 필수가 되었다. 따라서 이런 일들을 가능하게 해주는 수학지식을 설명하는 역할이 수학자들에게 주어졌다. 이에 본 연구진은 인공지능과 머신러닝(Machine Learning, 기계학습)을 이해하기 위해 필요한 수학 개념을 우리의 실정에 맞게 한 학기(또는 두 학기) 분량으로 정리하여, 무료 전자교과서 "인공지능을 위한 기초수학"을 집필하고, 인공지능 분야에 관심이 있는 다양한 전공의 대학생과 대학원생을 대상으로 하는 강좌를 개설하였다. 본 논문에서는 그 개발과정과 운영사례를 공유한다. http://matrix.skku.ac.kr/math4ai/