• Annual Conference of KIPS
• /
• 2024.05a
• /
• pp.711-714
• /
• 2024

In the dynamic landscape of modern machine learning, Federated Learning (FL) has emerged as a compelling paradigm designed to enhance privacy by enabling participants to collaboratively train models without sharing their private data. Specifically, Distillation-based Federated Learning, like Federated Learning with Model Distillation (FedMD), Federated Gradient Encryption and Model Sharing (FedGEMS), and Differentially Secure Federated Learning (DS-FL), has arisen as a novel approach aimed at addressing Non-IID data challenges by leveraging Federated Learning. These methods refine the standard FL framework by distilling insights from public dataset predictions, securing data transmissions through gradient encryption, and applying differential privacy to mask individual contributions. Despite these innovations, our survey identifies persistent vulnerabilities, particularly concerning the susceptibility to logit inversion attacks where malicious actors could reconstruct private data from shared public predictions. This exploration reveals that even advanced Distillation-based Federated Learning systems harbor significant privacy risks, challenging the prevailing assumptions about their security and underscoring the need for continued advancements in secure Federated Learning methodologies.

• Land and Housing Review
• /
• v.13 no.1
• /
• pp.11-22
• /
• 2022

This paper estimates the demand of newly married couples for public rental housing in Chungnam. This research attempts to overcome data limitations by linking survey data with administrative data for analysis. First, the results of a binary logit model that analyzes newly married couples' intention to move into public rental housing, based on the Chungnam Social Survey 2019, reveal that residential location, educational level, housing type, and tenure type have a statistically significant effect. By combining the estimated coefficients with another dataset, the statistics of newly married couples for administration purposes acquired from Statistics Korea, this research estimates the demand for public rental housing among the newly married couples in Chungnam. The estimation results show that the total demand for public rental housing in Chungnam is 11,424 units among 43,705 newly married couples. The total demand of 21,685 newly married couples who occupy rental housing is estimated to be 9,436 units. The policy for providing public rental housing to newly married couples in Chungnam aims to increase their fertility rates. Hence, further research should be followed up to evaluate the effect of the supply of public rental housing on fertility rates. Also, a research method should be developed to control for possible endogeneity between the demand for public rental housing and childbirths.

• Journal of Korea Technology Innovation Society
• /
• v.20 no.3
• /
• pp.546-575
• /
• 2017

This paper investigates the effects of public R&D support on medium-sized firms' R&D investment. The paper collects a panel dataset of Korean manufacturing firms' R&D investment and public support, and employs the DID (difference-in-differences) regression for the test of stimulating or crowding-out effect. Empirical analysis examines how the effect of public R&D support differs between small and medium-sized firms and whether firm size and technological capability moderate the effect in the sample of medium-sized firms. Empirical results show that public R&D support tends to generally stimulate private pure R&D investment for both small and medium-sized firms. Comparing the results for small and medium-sized firms, this paper finds that the stimulating effect is relatively larger and more significant for medium-sized firms, while the effect is not significant for small ones. Furthermore, the paper shows that the stimulating effect of public R&D subsidy on private R&D investment is relatively stronger for medium-sized firms with superior technological competence and the effect of tax support is greater for incompetent firms. These results suggest that public R&D policies and R&D programs, differentiated from those for existing small firms, are necessary for medium-sized firms to stimulate private R&D continuously and formulated carefully by considering firm size, technological capability and growth potential.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.11
• /
• pp.2690-2701
• /
• 2013

Humans are able to recognize small number of people they know well by the way they walk. This ability represents basic motivation for using human gait as the means for biometric identification. Such biometric can be captured at public places from a distance without subject's collaboration, awareness or even consent. Although current approaches give encouraging results, we are still far from effective use in practical applications. In general, methods set various constraints to circumvent the influence factors like changes of view, walking speed, capture environment, clothing, footwear, object carrying, that have negative impact on recognition results. In this paper we investigate the influence of walking speed variation to different visual based gait recognition approaches and propose normalization based on geometric transformations, which mitigates its influence on recognition results. With the evaluation on MoBo gait dataset we demonstrate the benefits of using such normalization in combination with different types of gait recognition approaches.

• New & Renewable Energy
• /
• v.6 no.2
• /
• pp.12-18
• /
• 2010

This paper reviews commercial wind mapping services provided by the oversea consulting companies, AL-PRO and anemos in Germany, AWS Truepower and 3TIER in USA. They provide quick-to-use but essential dataset for a preliminary assessment before commencing an actual feasibility study for wind farm development. Details of wind mapping method, mapresolution, data extraction height, price and so forth are compared and fresh service contents such as site analysis report are drawn from the comparison. Despite its public service, the objective value of the Renewable Energy Resource Map System of Korea Instistute of Energy Research is also confirmed and it is anticipated that the drawn new content idea will be ported to the system to enrich its applicability.

• Management Science and Financial Engineering
• /
• v.11 no.3
• /
• pp.137-153
• /
• 2005

Collaborative filtering-based recommendation systems make personalized recommendations based on users' ratings on products. Recommender systems must collect sufficient rating information from users to provide relevant recommendations because less user rating information results in poorer performance of recommender systems. To learn about new users, recommendation systems must first present users with an initial item list. In this study, we designed and analyzed seven selection strategies including the popularity, favorite, clustering, genre, and entropy methods. We investigated how these strategies performed using MovieLens, a public dataset. While the favorite and popularity methods tended to produce the highest average score and greatest average number of ratings, respectively, a hybrid of both favorite and popularity methods or a hybrid of demographic, favorite, and popularity methods also performed within acceptable ranges for both rating scores and numbers of ratings.

• YAKHAK HOEJI
• /
• v.56 no.1
• /
• pp.66-69
• /
• 2012

This study aims to assess the aggregate contribution of new drugs to the increase in life expectancy. We constructed a panel data combining mortality data in KOSIS and a drug dataset generated by assigning new drugs listed in 2000~2009 to their respective ICD codes. We found that 10% increase in stock of new drug led to 0.13~0.27% increase in the probability of survival to age 65. Due to lack of disease-specific life table, we used indirect approach to estimate the effect of new drugs on longevity. Using ordinary least squares, the estimate of the probability of survival to age 65 (logarithm) on life expectancy for all ages was 24.92. In conclusion, the increase in life expectancy of the entire population in Korea between 2000 and 2009 resulting from NMEs is 1.95 years, which explains 46.6% of real increase in life expectancy.

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.1
• /
• pp.9-16
• /
• 2016

In this paper, we present a performance evaluation agent based on fuzzy cluster analysis and validity measures. The proposed agent is consists of three modules, fuzzy cluster analyzer, performance evaluation measures, and feature ranking algorithm for feature selection step in CAD system. Feature selection is an important step commonly used to create more accurate system to help human experts. Through this agent, we get the feature ranking on the dataset of mass and calcification lesions extracted from the public real world mammogram database DDSM. Also we design a CAD system incorporating the agent and apply five different feature combinations to the system. Experimental results proposed approach has higher classification accuracy and shows the feasibility as a diagnosis supporting tool.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.28 no.1
• /
• pp.117-123
• /
• 2018

Objectives: The aim of this study was to examine factors influencing health-related quality of life in women workers using the dataset of the Korean National Health and Nutritional Examination Survey(KNHANES 2th). There were 955 subjects. Methods: A multiple regression model was used to study the factors influencing health-related quality of life of women workers. Results: A positive relationship was found between education(b=0.014, p=0.029) and health-related quality of life in women workers and non-osteoarthritis(b=0.037, p<0.001) and health-related quality of life in women workers. Conclusions: The results of this study show the importance of improving the working environment and preventing osteoarthritis in non-regular employment.

• Journal of Korea Multimedia Society
• /
• v.18 no.1
• /
• pp.1-8
• /
• 2015

In recent years, ear has emerged as a new biometric trait, because it has advantage of higher user acceptance than fingerprint and can be captured at remote distance in an indoor or outdoor environment. This paper proposes an individual identification method using ear region based on SIFT(shift invariant feature transform). Unlike most of the previous studies using rectangle shape for extracting a region of interest(ROI), this study sets an ROI as a flexible expanded region including ear. It also presents an effective extraction and matching method for SIFT keypoints. Experiments for evaluating the performance of the proposed method were performed on IITD public database. It showed correct identification rate of 98.89%, and it showed 98.44% with a deformed dataset of 20% occlusion. These results show that the proposed method is effective in ear recognition and robust to occlusion.