%66%69%6e%61%6e%63%69%61%6c
-
KAIST-KakaoBank Speeds Up 'Explainable AI' by 11 Times: "Boosts Financial AI Reliability
< (From left) Professor Jaesik Choi of the Kim Jaechul Graduate School of AI, Ph.D candidate Chanwoo Lee, Ph.D candidate Youngjin Park >
The research team led by Professor Jaesik Choi of KAIST's Kim Jaechul Graduate School of AI, in collaboration with KakaoBank Corp, announced that they have developed an accelerated explanation technology that can explain the basis of an Artificial Intelligence (AI) model's judgment in real-time. This research achievement significantly increases the practical applicability of Explainable Artificial Intelligence (hereinafter XAI) technology in fields requiring real-time decision-making, such as financial services, by achieving an average processing speed 8.5 times faster, and up to 11 times faster, than existing explanation algorithms for AI model predictions.
In the financial sector, a clear explanation for decisions made by AI systems is essential. Especially in services directly related to customer rights, such as loan screening and anomaly detection, regulatory demands to transparently present the basis for the AI model's judgment are increasingly stringent. However, conventional Explainable Artificial Intelligence (XAI) technologies required the repeated calculation of hundreds to thousands of baselines to generate accurate explanations, resulting in massive computational costs. This was a major factor limiting the application of XAI technology in real-time service environments.
To address this issue, Professor Choi's research team developed the 'ABSQR (Amortized Baseline Selection via Rank-Revealing QR)' framework for accelerating explanation algorithms. ABSQR noticed that the value function matrix generated during the AI model explanation process has a low-rank structure. It introduced a method to select only a critical few baselines from the hundreds available. This drastically reduced the computation complexity, which was previously proportional to the number of baselines, to be proportional only to the number of selected critical baselines, thereby maximizing computational efficiency while maintaining explanatory accuracy.
Specifically, ABSQR operates in two stages. The first stage systematically selects important baselines using Singular Value Decomposition (SVD) and Rank-Revealing QR decomposition techniques. Unlike existing random sampling methods, this is a deterministic selection method aimed at preserving information recovery, which guarantees the accuracy of the explanation while significantly reducing computation. The second stage introduces an amortized inference mechanism, which reuses the pre-calculated weights of the baselines through cluster-based search, allowing the system to provide an explanation for the model's prediction result in real-time service environments without repeatedly evaluating the model. The research team verified the superiority of ABSQR through experiments on various real-world datasets. Tests on standard datasets across five sectors—finance, marketing, and demographics—showed that ABSQR achieved an average processing speed 8.5 times faster than existing explanation algorithms that use all baselines, with a maximum speed improvement of over 11 times. Furthermore, the degradation of explanatory accuracy due to speed acceleration was minimized, maintaining up to 93.5% of the explanation accuracy compared to the baseline algorithm. This level is sufficient to meet the explanation quality required in real-world applications.
< ABSQR Framework Overview. (1) The baseline selection stage utilizes the low-rank structure of the value function matrix to select only a small number of key baselines, and (2) the accelerated search stage reuses the pre-calculated baseline weight coefficients based on clusters. This dramatically reduces the computation complexity, which was proportional to the number of baselines, to be proportional only to the number of selected key baselines. >
A KakaoBank official stated, "We will continue relentless research and development to enhance the reliability and convenience of financial services and introduce innovative financial technologies that customers can experience." Chanwoo Lee and Youngjin Park, co-first authors from KAIST, explained the significance of the research: "This methodology solves the crucial acceleration problem for real-time application in the financial sector, proving that it is possible to provide users with the reasons behind a learning model's decision in real-time." They added, "This research provides new insights into what constitutes unnecessary computation and the selection of important baselines in explanation algorithms, practically contributing to the improvement of explanation technology efficiency." This research, co-authored by PhD candidates Chanwoo Lee and Youngjin Park from the KAIST Kim Jaechul Graduate School of AI, and researchers Hyeongeun Lee and Yeeun Yoo from the KakaoBank Financial Technology Research Institute, was presented on November 12 at the 'CIKM 2025 (ACM International Conference on Information and Knowledge Management)', the world's highest-authority academic conference in the field of information and knowledge management. ※ Paper Title: Amortized Baseline Selection via Rank-Revealing QR for Efficient Model Explanation
※ Author Information:
※ Author Information: DOI: https://doi.org/10.1145/3746252.3761036
Co-First Authors: Chanwoo Lee (KAIST Kim Jaechul Graduate School of AI), Youngjin Park (KAIST Kim Jaechul Graduate School of AI), Hyeogeun Lee (KakaoBank), Yeeun Yoo (KakaoBank)
Co-Authors: Daehee Han (KakaoBank), Junho Choi (KAIST Kim Jaechul Graduate School of AI), Kunhyung Kim (KAIST Kim Jaechul Graduate School of AI)
Corresponding Authors: Nari Kim (KAIST Kim Jaechul Graduate School of AI), Jaesik Choi (KAIST Kim Jaechul Graduate School of AI)
Meanwhile, this research achievement was conducted through KakaoBank's industry-academia research project 'Advanced Research on Explainable Artificial Intelligence Algorithms in the Financial Sector' and the Ministry of Science and ICT/Institute for Information & Communications Technology Planning and Evaluation (IITP) supported project 'Development of Explainable Artificial Intelligence Technology Providing Explainability in a Plug-and-Play Manner and Verification of Explanation Provision for AI Systems.'
2025.12.12 View 3984 -
KAIST to Host FinTech Conference 2018
KAIST will be hosting a FinTech Conference with Princeton University (USA), Tsinghua University (China), and EDHEC Business School (France) in Seoul from April 12 to 13 titled State of the Art in Robo-Advising Systems: Financial Technologies for Enhanced Social Security.
The Bendhein Center for Finance and Department of Operations Research and Financial Engineering from Princeton University, Fintech Lab from Tsinghua University, and the Risk Institute from EDHEC will participate as co-hosting organizations in this conference organized by the Department of Industrial & Systems Engineering and the KAIST Center for Wealth Management Technologies.
The conference will discuss issues required for providing customized asset management to the public in terms of theory, technology, and industry.
During the conference, KAIST President Sung-Chul Shin and Chairman and CEO of the National Pension Service Sung Joo Kim will deliver welcoming addresses.
Professor John Mulvey from Princeton University, Professor Michael Dempster from Cambridge University, Professors Wei Xu and Changle Lin from Tsinghua University, Professor Lionel Martellini from EDHEC, and Professor Woo Chang Kim from KAIST are some of the invited speakers at the event.
Moreover, renowned experts in related fields will also participate in the conference, including Founder of Vanguard Group John Bogle, Jin Lee from Ant Financial, Youngsuh Cho from Shinhan Financial Group, Jung-Hwan Lee from Samsung Asset Management, and Hye Young Sung from the National Pension Service Research Institute.
Professor Kim said, “Only a small number of wealthy people can receive life-cycle customized asset management services due to the high cost structure; however, new technology derived from the Fourth Industrial Revolution can reduce the service price in an innovative manner, ultimately leading it to be popularized.”
“In an era with the poverty rate of older people reaching almost 50%.Fintech can enable individuals to manage their assets in an active manner, reinforcing social security without additional social costs in the period,” he added.
These four universities have been hosting the FinTech Conference since 2017. China will host the conference this fall, followed by France next year.
Samsung Asset Management, Alibaba Group, and Ant Financial will sponsor the conference. Anyone interested in this event can find more details at http://wmt.kaist.ac.kr/conference.html.
2018.04.02 View 13378 -
KAIST to Host Conference on the Financial Technology Industry
To take place in the Korea Federation of Banks building on May 21, 2015, Seoul
Around 200 individuals to be present including Oh-Kyu Kwon, the former Minister of Finance and Economy and Chang-Hyun Yun, the former president of Korea Institute of Finance
KAIST will hold a conference to discuss prospects for Korea’s financial technology industry (FinTech) and to address the rapid changes taking place in both national and global finance markets.
The conference will be held in the Convention Hall of the Korea Federation of Banks building in Myeongdong, Seoul on May 21, 2015 with a theme entitled “The Challenge and the Task of Korean FinTech Industry in the Global Finance Market.”
Around 200 individuals including Oh-Kyu Kwon, the former Minister of Finance and Economy, Jiwon Jung, a member of a standing committee of Financial Services Commission, and Steve Kang, the President of KAIST will attend the meeting.
The term “FinTech” originates from the combination of two words, finance and technology, to describe a new form of financial system based on Information Technology (IT). The precise definition and boundary of FinTech is not fully established yet, but this will provide Korea with an opportunity to take the lead in the finance industry with its cutting-edge IT.
The conference will focus on the methods by which Korean FinTech companies can acquire a competitive edge in the global finance market. The keynote speaker, Tong-Suk Kim, the Dean of College of Business, KAIST, will speak about “The Challenge and the Task of Korean FinTech.” The talk will emphasize how the FinTech industry can create an innovative industry, destroying the existing paradigm of payment system and Internet-only banks.
The keynote speech will be followed by a talk on “The Requirements for Establishing Korean FinTech Ecosystem” by Jong-Hyun Kim, a researcher at Woori Finance Research Institute, Korea.
Additional talks, which will be given by KAIST professors, are:
“Korean Economy and the Role of FinTech” by Professor Young-Sun Kwon of the Department Business and Technology Management;
“Korean FinTech in ICT Perspective” by Professor Yoon-Joon Lee of the Department of Computer Science;
“The Current State and Possibilities of Korean FinTech” by Professor Byung-Tae Lee of the School of Management Engineering; and
“The Task and Challenge of Promoting FinTech Industry” by Professor Byung-Chun Kim of the School of Management Engineering.
Also, Professor Chang-Hyun Yoon of the School of Business Administration, University of Seoul, will give a talk on “The Crisis in Korean Finance Industry and the Role of FinTech.”
The conference will also hold a discussion session on “The Promotion Methods for Korean FinTech Industry.” Participants for the discussion will include Jae-Sung Song from the Ministry of Science, ICT and Future Planning, Dong-Hwan Kim from the Financial Services Commission, Jeon Jin from Samsung Economic Research Institute (SERI), Hyo-Jin Lee of 8Percent, In-Chang Hwang from the Korea Insurance Research Institute, and Jae-Kyu Lee from Koscom.
Byung-Chun Kim, the Director of KAIST’s Financial Engineering Research Center, who organized the FinTech conference said, “FinTech industry will evolve from simply attempting to attract consumers to understanding the lifestyle patterns of consumers to provide customized service and information.”
2015.05.20 View 15635 -
KAIST College of Business ranks no.1 in Asia for executive education
KAIST College of Business ranked 28th in the world and 1st in Asia in the UK Financial Times (FT) 2012 Executive Education open ranking, making it the only school in Korea to be included.
The FT ranking selects the world’s top 65 business graduate schools based on executive education course design, teaching methods & materials, faculty, new skills & learning, quality of participants, and several other criteria. Last year, the KAIST executive education program ranked 29th in the world and 2nd in Asia. This year, it surpassed China’s CEIBS to rank 1st in Asia and has now been included in the ranking for three consecutive years.
Looking at the individual criteria, KAIST ranked highly in course preparation (11th), international location (11th), facilities (12th), and follow-up (5th) and saw increases in aims achieved (31st) and course design (29th).
KAIST attributed its success to the differentiation and specialization of education courses based on position, industry, and enterprise. KAIST College of Business operates an integrative course that fuses business management with science & technology, information media, medicine and innovation. The school runs both a chief and junior executive program to cater to different positional needs and provides specialized lectures that consider a company’s industry and size.
The ‘international location’ criteria saw a huge leap from 23rd to 11th due to the use of both long term and short term overseas exchange programs. The ‘Global Leader’ course created in 2011 is taught only in English and allows students to study abroad in a partner school in the second semester. The college’s knowhow in the MBA program also achieved recognition with high rankings in new skills & learning (18th) as well as in teaching methods & materials (24th).
College of Business President Lee Byung Tae said that the school’s success was achieved through a specialized curriculum that considers the market and explained that the ranking reflects the school’s competitiveness. He also said that KAIST will continue to provide the best educational services appropriate for KAIST’s outstanding reputation.
In the overall ranking, Swiss IMD was ranked number 1, followed by the Harvard Business School. In Asia, China’s CEIBS (29th), Singapore’s Aalto University (42nd), and the National University of Singapore Business School (64th) followed suit.
2012.05.17 View 15180 -
KAIST MBA program was ranked 99th by Financial Times Global MBA Rankings 2011.
The Business School of KAIST in Seoul made the 99th placement among the top 100 business schools in the world, according to the Financial Times Global MBA Rankings 2011. The Hindustan Times, an English daily newspaper in India, brifely carried an article on this topic. For the article, please go to http://www.hindustantimes.com/StoryPage/Print/656953.aspx.
Financial Times Business Education and School Rakings 2011: http://www.ft.com/businesseducation/mba2011
2011.02.03 View 14780
-
KAIST Signs MOU with Macquarie for Cooperation in Green Growth Projects
KAIST and Shinhan Macquarie Financial Advisory Co. have reached an agreement for cooperation in the development and commercialization of the KAIST-led two national green growth projects, On-line Electric Vehicle (OLEV) and Mobile Harbor (MH) programs, university authorities said.
KAIST President Nam-Pyo Suh signed a memorandum of understanding with John Walker, Chairman of Macquarie Group of Companies in Korea, and Woo-Gon Hwang, Representative Director of Shinhan Macquarie Financial Advisory Co. on Wednesday, Oct. 21.
Under the agreement, KAIST and Macquarie will cooperate in developing an optimal business structure for a rapid commercialization of OLEV and MH. Specifically, Shinhan Macquarie Financial Advisory will provide financial advice, including basis financial analysis, potential investor inducement and feasibility analysis of the projects.
Shinhan Macquarie Financial Advisory Co. is a joint venture between Shinhan Financial Group of Korea and the Australia-based Macquarie Bank Group which provides global investment banking and diversified financial services.
KAIST"s OLEV is a project to develop a new growth engine for Korea and lead the future of global automotive industry. It is an entirely new concept: the electric vehicle picks up power from underground power supplier lines, while either running or standing, through the non-contact magnetic charging method.
The MH program is designed to develop a system that can load/unload containers from a containership in the open sea and deliver them to their destinations at the harbor.
The Korean government has included these KAIST projects, which both are great technical and engineering challenges, in the nation"s sustainable growth programs, providing substantial research grants. KAIST offers its advanced research capabilities for the nation"s efforts to achieve efficient, environment-friendly utilization of resources as new growth engines that spur the development of related industries and explore global markets.
2009.10.22 View 21327 -
World Research University Heads to Discuss Challenges in Global Financial Turmoil
About 70 leaders of the world"s major research universities will discuss how to better contribute to continued development of human society in global financial turmoil at a symposium organized by KAIST Monday (Sept. 21) at the Westin Chosun Hotel in Seoul.
Participants of the 2nd International Presidential Forum on Global Research Universities are from 40 universities in 25 countries, including Stanford University and Georgia Institute of Technology of the United States, Berlin Institute of Technology of Germany, Paris Institute of Technology of France, Technical University of Denmark, National University of Singapore and Tokyo Institute of Technology. They include 20 presidents of Korean universities and two dozens of leaders from industry and the government.
Under the main subject of "Challenges to Global Research Universities," the international symposium will proceed in four panel sessions. The subjects of each session and their keynote speakers are:
-- "Institutional Management in Times of Financial Crisis" by Kurt Kutzler, President of Berlin Institute of Technology
-- "Innovations in Education & Research" by Brian Cantor, Vice Chancellor of University of York
-- "Globalization of Institutes of Higher Learning" by Gary Schuster, Provost and Executive Vice President of Georgia Institute of Technology
-- "The Roles of Government, University and Industry in Green Technology Development" by KAIST President Nam-Pyo Suh
KAIST President Suh said of the purpose of the conference: "The world has witnessed a global financial turmoil of unseen magnitude and many nations are still struggling under the devastating impacts. While universities were no exception in facing economic turmoil, they have realized renewed pressures and expectations from their respective communities to provide answers to the great challenges."
"The conference will serve as an opportunity for the representatives of research universities to compare their visions of networking among theier institutions and initiate steps for new relationships. The conference I am sure will have a far-reaching influence on the course our research universities will take to shoulder greater responsibilities for building a better future of the mankind."
For more information, visit forum.kaist.ac.kr
2009.09.16 View 27698