KOINEU Logo
No Image

Fragile $mathit{vs}$ robust Multiple Equilibria phases in generalized Lotka-Volterra model with non-reciprocal interactions

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ๋น„๋Œ€์นญ ์ƒํ˜ธ์ž‘์šฉ ์€ ์‹ ๊ฒฝ๋ง, ๋ฏธ์ƒ๋ฌผ ๊ตฐ์ง‘, ์—ด๋Œ€์šฐ๋ฆผ ๋“ฑ ์ˆ˜๋งŽ์€ ์ด์งˆ์  ๊ตฌ์„ฑ์š”์†Œ๊ฐ€ ๋น„๋Œ€์นญ์ ์œผ๋กœ ์–ฝํžŒ ์‹œ์Šคํ…œ์—์„œ ํ•ต์‹ฌ์ ์ธ ์—ญํ• ์„ ํ•œ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ(R. May ๋“ฑ)๋Š” ๋ฌด์ž‘์œ„ ์ƒํ˜ธ์ž‘์šฉ ์„ ๊ฐ€์ •ํ•˜๊ณ  ๋‹ค์–‘์„ฑโ€‘์•ˆ์ •์„ฑ ๋ฌธ์ œ๋ฅผ ๋‹ค๋ฃจ์—ˆ์ง€๋งŒ, ๋น„๋Œ€์นญ์„ฑ(ฮณ < 1)์ด ๋„์ž…๋˜๋ฉด ํ˜ผ๋ˆยท๋น„ํ‰ํ˜• ํ˜„์ƒ ์ด ๋‚˜ํƒ€๋‚œ๋‹ค. ๋”ฐ๋ผ์„œ ๋น„๋Œ€์นญ์„ฑ์„ ํฌํ•จํ•œ ์ผ๋ฐ˜ํ™”๋œ ๋กœํŠธ์นดโ€‘๋ณผํ…Œ๋ผ(gLV) ๋ชจ๋ธ ์—์„œ ํ‰ํ˜•์ ๋“ค์˜ ์ˆ˜์™€ ํŠน์„ฑ์„ ์ •๋Ÿ‰ํ™”ํ•˜๋Š” ๊ฒƒ์ด ํ•„์š”ํ•˜๋‹ค. 2. ๋ชจ๋ธ ์ •์˜ ์ข… ํ’๋ถ€๋„ (N i) (i 1โ€ฆS) ์˜ ๋™์—ญํ•™์€

Condensed Matter Model
No Image

Fragment-Based Configuration Interaction: Towards a Unifying Description of Biexcitonic Processes in Molecular Aggregates

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๋น„์—‘์‹œํ†ค์˜ ๋‹ค๋ฉด์„ฑ : ์‹ฑ๊ธ€๋ › ํŒŒ์ „, ํŠธ๋ฆฌํ”Œ๋ ›โ€‘ํŠธ๋ฆฌํ”Œ๋ › ์†Œ๋ฉธ, ๊ณ ์—๋„ˆ์ง€ ์ „ํ•˜ ์ƒ์„ฑ ๋“ฑ ๋‹ค์–‘ํ•œ ํ˜„์ƒ์— ๊ด€์—ฌํ•˜์ง€๋งŒ, ๊ธฐ์กด ์ด๋ก ์€ ์ฃผ๋กœ ๋‹จ์ผโ€‘์—‘์‹œํ†ค (TDโ€‘DFT, LRโ€‘CC) ํ˜น์€ ํŠน์ • ๋น„์—‘์‹œํ†ค (์˜ˆ: TT)๋งŒ์„ ๋‹ค๋ฃจ์–ด ์™”๋‹ค. ํ†ตํ•ฉ ํ”„๋ ˆ์ž„์›Œํฌ ๋ถ€์žฌ : LELE, CTCT, TT, ๊ทธ๋ฆฌ๊ณ  ํ˜ผํ•ฉ CTX์™€ ๊ฐ™์€ ๋ชจ๋“  ๋‘โ€‘์ž…์ž ๊ตฌ์„ฑ์š”์†Œ๋ฅผ ๋™์‹œ์— ๋‹ค๋ฃจ๋Š” ํ™”ํ•™์ ์œผ๋กœ ํ•ด์„ ๊ฐ€๋Šฅํ•œ ๋ชจ๋ธ์ด ์—†์—ˆ๋‹ค. ์ด๋Š” ํŠนํžˆ ๋‹ค์ค‘โ€‘ํฌ๋ฆฌ์Šคํ„ธยท๋Œ€๊ทœ๋ชจ ์ง‘ํ•ฉ์ฒด ์—์„œ ์ค‘์š”ํ•œ๋ฐ, ์‹คํ—˜์  ๊ด€์ธก(2D ์ „์ž ์ŠคํŽ™ํŠธ๋กœ์Šค์ฝ”ํ”ผ, ๊ณ ๋ฐ€๋„ ํŽŒํ”„โ€‘ํ”„๋กœ๋ธŒ ๋“ฑ)๊ณผ ์ด๋ก  ์‚ฌ

Physics
No Image

Free Lunch in Medical Image Foundation Model Pre-training via Randomized Synthesis and Disentanglement

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ๋ฐ์ดํ„ฐ ๋ณ‘๋ชฉ : ์˜๋ฃŒ ์˜์ƒ์€ ์ดฌ์˜ ๋น„์šฉ, ๊ฐœ์ธ์ •๋ณด ๊ทœ์ œ, ์žฅ๋น„ยท๊ธฐ๊ด€ ๊ฐ„ ์ด์งˆ์„ฑ ๋“ฑ์œผ๋กœ ๋Œ€๊ทœ๋ชจ ๋ผ๋ฒจ๋ง์ด ์–ด๋ ต๋‹ค. ๊ธฐ์กด ์ ‘๊ทผ๋ฒ• ํ•œ๊ณ„ : ์ง€๋„ ํ•™์Šต โ†’ ๋ผ๋ฒจ ๋น„์šฉ ํญ์ฆ. ์ž๊ธฐ์ง€๋„ ํ•™์Šต โ†’ ๊ฐ•๋ ฅํ•œ ์‹ ํ˜ธ ๋ถ€์กฑ, ๋ณตํ•ฉ ์˜๋ฏธ disentanglement ์–ด๋ ค์›€. ์ƒ์„ฑ ๊ธฐ๋ฐ˜ ํ•ฉ์„ฑ โ†’ ์‹ค์ œ ๋ฐ์ดํ„ฐ์— ํฌ๊ฒŒ ์˜์กด, ๋น„์šฉยทํ”„๋ผ์ด๋ฒ„์‹œ ๋ฌธ์ œ ์™„์ „ ํ•ด์†Œ ์•ˆ ๋จ. ํ•ต์‹ฌ ๊ฐ€์„ค : โ€œํ•ฉ์„ฑ ์˜๋ฃŒ ์˜์ƒ๋งŒ์œผ๋กœ๋„ ๊ธฐ์ดˆ ๋ชจ๋ธ์„ ์ถฉ๋ถ„ํžˆ ์‚ฌ์ „ํ•™์Šตํ•  ์ˆ˜ ์žˆ๋‹ค.โ€ 2. RaSD ํ•ต์‹ฌ ์•„์ด๋””์–ด | ์š”์†Œ | ๊ตฌํ˜„ ๋ฐฉ์‹ | ๊ธฐ๋Œ€ ํšจ๊ณผ | | | | | | ๊ตฌ์กฐ

Quantitative Biology Model
No Image

From Chain-Ladder to Individual Claims Reserving

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ 1. ๊ฐœ๋ณ„ ์ฒญ๊ตฌ ์ˆ˜์ค€ ์˜ˆ์ธก์˜ ํ˜„์ฃผ์†Œ ์ง€๋‚œ 10๋…„๊ฐ„ ํšŒ๊ท€ ํŠธ๋ฆฌ, ๊ทธ๋ž˜๋””์–ธํŠธ ๋ถ€์ŠคํŒ…, ์‹ ๊ฒฝ๋ง ๋“ฑ ๋‹ค์–‘ํ•œ ML ๊ธฐ๋ฒ•์ด ์ œ์•ˆ๋์ง€๋งŒ, ์‚ฐ์—… ํ˜„์žฅ์—์„œ๋Š” ์•„์ง ํ‘œ์ค€ํ™”๋œ ๋ฐฉ๋ฒ•์ด ๋ถ€์žฌํ•จ. ์ฃผ์š” ์žฅ์•  ์š”์ธ: ๊ณต๊ฐœ ๋ฐ์ดํ„ฐ ๋ถ€์กฑ, ๋ฐ์ดํ„ฐ์˜ ๊ฒ€์—ดยท์ €๋นˆ๋„ยท๋ณตํ•ฉ ์‹œ๊ณ„์—ด ํŠน์„ฑ, ๋‹ค๊ธฐ๊ฐ„ ์˜ˆ์ธก์„ ์œ„ํ•œ ๋ชจ๋ธ ์„ค๊ณ„ ์–ด๋ ค์›€, ๊ตฌํ˜„ ๋ณต์žก์„ฑ ๋ฐ ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ํŠœ๋‹ ๋ถ€๋‹ด. 2. ์ „ํ†ต์  CL ๋ฐฉ๋ฒ•์˜ ์žฅ์  ์ง‘๊ณ„ ์ˆ˜์ค€์—์„œ ๋†’์€ ์ •ํ™•๋„์™€ ํ•ด์„ ์šฉ์ด์„ฑ, ์ˆ˜์‹ญ ๋…„๊ฐ„ ๊ฒ€์ฆ๋œ ์‹ค๋ฌด ์ ์šฉ์„ฑ. ๊ทธ๋Ÿฌ๋‚˜ ๊ฐœ๋ณ„ ์ฒญ๊ตฌ์— ์ง์ ‘ ์ ์šฉํ•˜๊ธฐ์—” โ€œ๋‹จ๊ณ„๋ณ„ ์ „์ง„โ€ ๋ฐฉ์‹์ด ๋‹ค๊ธฐ๊ฐ„ ์˜ˆ์ธก์—

Statistics Applications
From Verification Burden to Trusted Collaboration: Design Goals for LLM-Assisted Literature Reviews

From Verification Burden to Trusted Collaboration: Design Goals for LLM-Assisted Literature Reviews

๋ณธ ๋…ผ๋ฌธ์€ ๋Œ€ํ˜• ์–ธ์–ด ๋ชจ๋ธ(LLMs)์ด ํ•™์ˆ  ๊ธ€์“ฐ๊ธฐ์—์„œ ์–ด๋–ป๊ฒŒ ํ™œ์šฉ๋˜๋Š”์ง€์— ๋Œ€ํ•œ ๊นŠ์ด ์žˆ๋Š” ๋ถ„์„์„ ์ œ๊ณตํ•˜๋ฉฐ, ํŠนํžˆ ๋ฌธํ—Œ ๊ฒ€ํ†  ๊ณผ์ •์—์„œ์˜ ๊ตฌ์ฒด์ ์ธ ์ ์šฉ ๋ฐฉ๋ฒ•๊ณผ ์ œํ•œ ์‚ฌํ•ญ์„ ํƒ์ƒ‰ํ•ฉ๋‹ˆ๋‹ค. ์—ฐ๊ตฌ์ž๋“ค์ด LLMs๋ฅผ ์‚ฌ์šฉํ•˜๋ฉด์„œ ๊ฒช๋Š” ์ฃผ์š” ๋ฌธ์ œ๋กœ ์‹ ๋ขฐ์„ฑ ๋ถ€์กฑ, ์ง€์†์ ์ธ ๊ฒ€์ฆ ํ•„์š”์„ฑ, ๊ทธ๋ฆฌ๊ณ  ์—ฌ๋Ÿฌ ๋„๊ตฌ์˜ ๋ณต์žก์„ฑ์„ ๋“ค ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ๋ฌธ์ œ์ ๋“ค์„ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด ๋…ผ๋ฌธ์€ 6๊ฐ€์ง€ ์„ค๊ณ„ ๋ชฉํ‘œ์™€ ์ด๋ฅผ ์‹คํ˜„ํ•˜๋Š” ๊ณ ์ˆ˜์ค€ ํ”„๋ ˆ์ž„์›Œํฌ๋ฅผ ์ œ์•ˆํ•ฉ๋‹ˆ๋‹ค. ์ด ํ”„๋ ˆ์ž„์›Œํฌ๋Š” ๊ด€๋ จ ๋…ผ๋ฌธ์„ ์‹œ๊ฐํ™”ํ•˜์—ฌ ์‰ฝ๊ฒŒ ์ดํ•ดํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•˜๊ณ , ๊ฐ ๋‹จ๊ณ„์—์„œ์˜ ๊ฒ€์ฆ ๊ณผ์ •์„ ํ†ตํ•ด ์‹ ๋ขฐ์„ฑ์„ ๋†’์ด

No Image

Fully-automated sleep staging: multicenter validation of a generalizable deep neural network for Parkinson's disease and isolated REM sleep behavior disorder

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ iRBD โ†’ ฮฑโ€‘์‹œ๋ˆ„ํด๋ ˆ์ธ๋ณ‘ : iRBD๋Š” ํŒŒํ‚จ์ŠจยทDLBยทMSA ๋“ฑ ฮฑโ€‘์‹œ๋ˆ„ํด๋ ˆ์ธ๋ณ‘๊ตฐ์˜ ๊ฐ€์žฅ ๊ฐ•๋ ฅํ•œ ์ „๊ตฌํ‘œ์ง€์ด๋ฉฐ, 12โ€‘14๋…„ ์ถ”์  ์‹œ 73โ€‘90 %๊ฐ€ ์‹ค์ œ ์‹ ๊ฒฝํ‡ดํ–‰์„ฑ ์งˆํ™˜์œผ๋กœ ์ง„ํ–‰ํ•œ๋‹ค. vPSG์˜ ๋ณ‘๋ชฉ : ํ˜„์žฌ ์ง„๋‹จ ํ‘œ์ค€์ธ vPSG๋Š” ์ˆ˜๋™ 30 s epoch ๋‹จ์œ„ ์ˆ˜๋ฉด ๋‹จ๊ณ„ ์ฑ„์ ์ด ๋งค์šฐ ๋…ธ๋™์ง‘์•ฝ์ ์ด๋ฉฐ, ํŠนํžˆ PDยทiRBD์—์„œ EEGยทEOGยทEMG ๋ณ€ํ˜•์œผ๋กœ ์ธํ•ด ์ธ๊ฐ„ ์ฑ„์ ์ž์˜ ์ผ๊ด€์„ฑ์ด ๋–จ์–ด์ง„๋‹ค (ฮบ โ‰ˆ 0.61). ์ž๋™ํ™” ๊ธฐ๋Œ€ : ๋”ฅ๋Ÿฌ๋‹ ๊ธฐ๋ฐ˜ ์ž๋™ ์ˆ˜๋ฉด ๋‹จ๊ณ„ ๊ตฌ๋ถ„์€ ์ธ๊ฐ„ ์ฑ„์ ์ž์˜ ๋ณ€๋™์„ฑ์„ ๊ฐ์†Œ์‹œํ‚ค๊ณ , ๋Œ€๊ทœ

Computer Science Network Machine Learning
No Image

Functional Central Limit Theorem for Stochastic Gradient Descent

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ SGD์˜ ๋น„๋Œ€์นญ์  ๋ณ€๋™ : ๊ธฐ์กด ๋ฌธํ—Œ์€ ์ฃผ๋กœ ์ตœ์ข… ์ดํ„ฐ๋ ˆ์ดํŠธ ํ˜น์€ ํ‰๊ท ๊ฐ’์— ๋Œ€ํ•œ ์ wise CLT์— ์ง‘์ค‘ํ–ˆ์œผ๋ฉฐ, ์ „์ฒด ๊ฒฝ๋กœ์˜ ์‹œ๊ฐ„์  ์ƒ๊ด€๊ตฌ์กฐ๋Š” ๊ฑฐ์˜ ๋‹ค๋ฃจ์ง€ ์•Š์•˜๋‹ค. ๋น„์Šค๋ฌด์Šคยท๋น„๊ฐ•ํ•œ ๋ณผ๋ก ๋ฌธ์ œ : ๋กœ๋ฒ„์ŠคํŠธ ์œ„์น˜ ์ถ”์ •(์˜ˆ: ๊ธฐํ•˜ํ•™์  ์ค‘์•™๊ฐ’)์ฒ˜๋Ÿผ ๋ฏธ๋ถ„ ๊ฐ€๋Šฅ์„ฑ์ด ์—†๊ฑฐ๋‚˜ ์ „์—ญ ๊ฐ•ํ•œ ๋ณผ๋ก์„ฑ์ด ๊ฒฐ์—ฌ๋œ ๊ฒฝ์šฐ์—๋„ SGD๊ฐ€ ๋„๋ฆฌ ์‚ฌ์šฉ๋œ๋‹ค. ์ด๋Ÿฌํ•œ ์ƒํ™ฉ์—์„œ์˜ asymptotic behavior๋Š” ์•„์ง ๋ฏธ๋น„ํ–ˆ๋‹ค. 2. ์ฃผ์š” ๊ธฐ์—ฌ | ๋ฒˆํ˜ธ | ๋‚ด์šฉ | ์˜์˜ | | | | | | โ‘  | ํ•จ์ˆ˜ํ˜• ์ค‘์‹ฌ๊ทนํ•œ์ •๋ฆฌ(FCLT) ์ฆ๋ช… : ์žฌ์Šค์ผ€

Statistics Machine Learning
No Image

Functional Decomposition and Shapley Interactions for Interpreting Survival Models

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์ƒ์กด ๋ถ„์„์—์„œ ์ƒํ˜ธ์ž‘์šฉ ์€ ์ž„์ƒยท๊ณต์ค‘๋ณด๊ฑด ์˜์‚ฌ๊ฒฐ์ •์— ํ•ต์‹ฌ์ ์ด๋ฉฐ, ์œ ์ „ยทํ™˜๊ฒฝ, ๋น„๋งŒยท์น˜๋ฃŒ, ์—ฐ๋ นยท์ข…์–‘ํ‘œ์ง€์ž ๋“ฑ ๋‹ค์–‘ํ•œ ์‚ฌ๋ก€๊ฐ€ ์กด์žฌํ•œ๋‹ค. ๊ธฐ์กด ํ†ต๊ณ„์  ์ƒ์กด ๋ชจ๋ธ (CoxPH, Aalen ๋“ฑ)์€ ๋ช…์‹œ์  ๊ตํ˜ธํ•ญ์„ ํ†ตํ•ด ์ƒํ˜ธ์ž‘์šฉ์„ ํƒ์ƒ‰ํ•˜์ง€๋งŒ, ๋จธ์‹ ๋Ÿฌ๋‹ ๊ธฐ๋ฐ˜ ์ƒ์กด ๋ชจ๋ธ ์€ ๋ณต์žกํ•œ ๋น„์„ ํ˜•ยท์‹œ๊ฐ„โ€‘๋ณ€ํ™” ์ƒํ˜ธ์ž‘์šฉ์„ ์ž๋™์œผ๋กœ ํ•™์Šตํ•˜๋ฉด์„œ๋„ โ€œ๋ธ”๋ž™๋ฐ•์Šคโ€ ํŠน์„ฑ ๋•Œ๋ฌธ์— ํ•ด์„์ด ์–ด๋ ค์› ๋‹ค. Shapley ๊ธฐ๋ฐ˜ ํ•ด์„ (SurvLIME, SurvSHAP(t) ๋“ฑ)์€ ๊ฐœ๋ณ„ ํ”ผ์ฒ˜ ๊ธฐ์—ฌ๋Š” ์ œ๊ณตํ•˜์ง€๋งŒ ๊ณ ์ฐจ์› ์ƒํ˜ธ์ž‘์šฉ ์„ ๋‹ค๋ฃจ์ง€ ๋ชปํ•œ๋‹ค. 2. ํ•ต์‹ฌ

Statistics Machine Learning Model
No Image

GAMA: A Neural Neighborhood Search Method with Graph-aware Multi-modal Attention for Vehicle Routing Problem

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ VRP ๋Š” ๋ฌผ๋ฅ˜ยท๋ฐฐ๋‹ฌ ๋ถ„์•ผ์—์„œ ํ•ต์‹ฌ์ ์ธ NPโ€‘hard ๋ฌธ์ œ์ด๋ฉฐ, ์ „ํ†ต์ ์ธ ํœด๋ฆฌ์Šคํ‹ฑยท๋ฉ”ํƒ€ํœด๋ฆฌ์Šคํ‹ฑ์ด ๊ฐ•๋ ฅํ•˜์ง€๋งŒ ํŒŒ๋ผ๋ฏธํ„ฐ ํŠœ๋‹ยท๋„๋ฉ”์ธ ์ง€์‹์— ํฌ๊ฒŒ ์˜์กดํ•œ๋‹ค. Neural Neighborhood Search (NNS) ๋Š” ๊ฐ•ํ™”ํ•™์Šต ๊ธฐ๋ฐ˜ ์ •์ฑ…์ด ํ˜„์žฌ ํ•ด์˜ ์ด์›ƒ์„ ํƒ์ƒ‰ํ•˜๋„๋ก ํ•™์Šต์‹œ์ผœ, ๋น ๋ฅธ ๊ทผ์‚ฌํ•ด๋ฅผ ์ œ๊ณตํ•œ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๊ธฐ์กด NNS๋Š” 1) ๋‹จ์ผ ๋ชจ๋‹ฌ๋ฆฌํ‹ฐ (์˜ˆ: ๊ฒฝ๋กœ ์ˆœ์„œ๋งŒ)๋กœ ์ƒํƒœ๋ฅผ ํ‘œํ˜„, 2) ๋‹จ์ˆœ ์—ฐ๊ฒฐ ๋ฐฉ์‹์œผ๋กœ ์—ฌ๋Ÿฌ ์ •๋ณด๋ฅผ ๊ฒฐํ•ฉํ•ด ๊ตฌ์กฐ์  ์ •๋ณด๋ฅผ ์†์‹คํ•œ๋‹ค๋Š” ํ•œ๊ณ„๊ฐ€ ์žˆ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด โ€“ GAMA | ์š”์†Œ |

No Image

Generalised Exponential Kernels for Nonparametric Density Estimation

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๊ฒฝ๊ณ„ ํŽธํ–ฅ ๋ฌธ์ œ : ์ „ํ†ต์ ์ธ ๋Œ€์นญ ์ปค๋„์€ (mathbb{R}) ์ „์—ญ์„ ์ง€์›ํ•˜์ง€๋งŒ, ์–‘์˜ ๋ฐ์ดํ„ฐ์— ์ ์šฉํ•˜๋ฉด ์Œ์ˆ˜ ์˜์—ญ์— ๋น„์ •์ƒ์ ์ธ ํ™•๋ฅ  ์งˆ๋Ÿ‰์„ ํ• ๋‹นํ•ด ๊ฒฝ๊ณ„ ํŽธํ–ฅ์ด ๋ฐœ์ƒํ•œ๋‹ค. ๋น„๋Œ€์นญ ์ปค๋„์˜ ๋ถ€์ƒ : Chen(2000)์˜ ๊ฐ๋งˆ ์ปค๋„, Scaillet(2004)์˜ ์—ญ๊ฐ€์šฐ์‹œ์•ˆยท์—ญ์—ญ๊ฐ€์šฐ์‹œ์•ˆ ์ปค๋„ ๋“ฑ ๋‹ค์–‘ํ•œ ๋น„๋Œ€์นญ ์ปค๋„์ด ์ œ์•ˆ๋์ง€๋งŒ, ๋Œ€๋ถ€๋ถ„ ํŠน์ˆ˜ํ•จ์ˆ˜(๊ฐ๋งˆยท๋ฒ ํƒ€ ๋“ฑ)์— ์˜์กดํ•ด ๊ตฌํ˜„ยท๋ถ„์„์ด ๋ณต์žกํ•˜๋‹ค. GE ์ปค๋„์˜ ์ฐจ๋ณ„์  : GE ๋ถ„ํฌ๋Š” (alpha, lambda>0) ํŒŒ๋ผ๋ฏธํ„ฐ๋งŒ์œผ๋กœ ์ •์˜๋˜๋ฉฐ, (alpha>

Statistics
No Image

Generalised Mรถbius Categories and Convolution Kleene Algebras

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์ปจ๋ณผ๋ฃจ์…˜ ๋Œ€์ˆ˜ ๋Š” ํ•จ์ˆ˜(๋˜๋Š” ์‚ฌ์ƒ)๋“ค์˜ โ€œ๊ณฑ์…ˆโ€์„ ๊ตฌ์กฐ์  ํ•ฉ์„ฑ์œผ๋กœ ์ •์˜ํ•จ์œผ๋กœ์จ, ๋‹ค์–‘ํ•œ ์ˆ˜ํ•™ยท์ปดํ“จํ„ฐ ๊ณผํ•™ ๋ถ„์•ผ(์˜ˆ: ์‹ ํ˜ธ ์ฒ˜๋ฆฌ, ์ž๋™์ž ์ด๋ก , ํ”„๋กœ๊ทธ๋žจ ์˜๋ฏธ๋ก )์—์„œ ํ•ต์‹ฌ ๋„๊ตฌ๊ฐ€ ๋œ๋‹ค. ํด๋ ˆ์ธ ๋Œ€์ˆ˜ ๋Š” ๊ฐ€๋ฒ•์  ๋ฉฑ๋“ฑ์„ฑ์„ ๊ฐ–๋Š” ๋ฐ˜์ •์—ญ์— ๋ณ„ ์—ฐ์‚ฐ(โ˜…)์„ ์ถ”๊ฐ€ํ•œ ๊ตฌ์กฐ๋กœ, ์ •๊ทœ ํ‘œํ˜„์‹ยทํ”„๋กœ๊ทธ๋žจ ๊ฒ€์ฆ ๋“ฑ์—์„œ โ€œ๋ฐ˜๋ณตโ€์„ ๋ชจ๋ธ๋งํ•œ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ์—์„œ๋Š” ํŠน์ •ํ•œ ๊ตฌ์กฐ(์˜ˆ: ์ž์œ  ๋ชจ๋…ธ์ด๋“œ, ํŠธ๋ฆฌ, ๋ถ€๋ถ„ ์ˆœ์„œ ์ง‘ํ•ฉ) ์œ„์—๋งŒ ๋ณ„ ์—ฐ์‚ฐ์„ ์ •์˜ํ•ด ์™”์œผ๋ฉฐ, ๋ณด๋‹ค ์ผ๋ฐ˜์ ์ธ ๋ฒ”์ฃผ๋‚˜ ๊ณ ์ฐจ ๊ตฌ์กฐ์— ์ ์šฉํ•˜๊ธฐ๋Š” ์–ด๋ ค์› ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด 1. ์ผ

No Image

Generalized bilinear Koopman realization from input-output data for multi-step prediction with metaheuristic optimization of lifting function and its application to real-world industrial system

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์‚ฐ์—… ํ˜„์žฅ์˜ ์ œ์•ฝ : ์„ผ์„œ ์„ค์น˜ ๋น„์šฉยท๊ณต๊ฐ„ ์ œํ•œ์œผ๋กœ ์ „์ฒด ์ƒํƒœ๋ฅผ ์ธก์ •ํ•˜๊ธฐ ์–ด๋ ค์›€. Koopman์˜ ์žฅ์  : ๋น„์„ ํ˜•์„ ์„ ํ˜•์œผ๋กœ ๊ณ ์ฐจ์› ๋งคํ•‘ํ•ด ๊ธฐ์กด ์„ ํ˜• ์ œ์–ด ์ด๋ก  ์ ์šฉ ๊ฐ€๋Šฅ. ๊ธฐ์กด ํ•œ๊ณ„ LTI Koopman โ†’ ์ œ์–ดโ€‘๋น„์„ ํ˜• ์‹œ์Šคํ…œ์„ ์ถฉ๋ถ„ํžˆ ํฌ์ฐฉ ๋ชปํ•จ. ๋ฆฌํ”„ํŒ… ํ•จ์ˆ˜ ์„ค๊ณ„๊ฐ€ ์ˆ˜์ž‘์—…์— ์˜์กด, ์žฌํ˜„์„ฑยท์„ฑ๋Šฅ ์ €ํ•˜. NN ๊ธฐ๋ฐ˜ ๋ฆฌํ”„ํŒ…์€ ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ํŠœ๋‹ยท๊ณผ์ ํ•ฉยทํ•™์Šต ๋น„์šฉ ๋ฌธ์ œ. 2. ์ฃผ์š” ๊ธฐ์—ฌ | ๋ฒˆํ˜ธ | ๊ธฐ์—ฌ ๋‚ด์šฉ | ๊ธฐ์กด ์—ฐ๊ตฌ์™€ ์ฐจ๋ณ„์  | | | | | | โ‘  | ์ž…ยท์ถœ๋ ฅ ์ „์šฉ ์ด์ค‘์„ ํ˜• Koopman ์‹คํ˜„ : ์‹œ

Data System Electrical Engineering and Systems Science
No Image

Generalized Leverage Score for Scalable Assessment of Privacy Vulnerability

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ํ˜„๋Œ€ ๋”ฅ๋Ÿฌ๋‹ ๋ชจ๋ธ์€ ํ•™์Šต ๋ฐ์ดํ„ฐ์˜ ์ผ๋ถ€๋ฅผ ๊ธฐ์–ต(memorization) ํ•˜๋ฉฐ, ์ด๋Š” Membership Inference Attack(MIA)์ด๋ผ๋Š” ํ”„๋ผ์ด๋ฒ„์‹œ ์œ„ํ˜‘์„ ์ดˆ๋ž˜ํ•œ๋‹ค. ๊ธฐ์กด ๊ฐœ๋ณ„ ์œ„ํ—˜ ํ‰๊ฐ€ ๋ฐฉ๋ฒ•์€ ๊ทธ๋ฆผ์ž ๋ชจ๋ธ ์„ ๋‹ค์ˆ˜ ํ•™์Šต์‹œ์ผœ์•ผ ํ•˜๋ฏ€๋กœ ๋Œ€๊ทœ๋ชจ ๋ชจ๋ธยท๋ฐ์ดํ„ฐ์…‹์— ์ ์šฉํ•˜๊ธฐ ์–ด๋ ต๋‹ค. ๋”ฐ๋ผ์„œ โ€œ ์žฌํ•™์Šต ์—†์ด ๊ฐœ๋ณ„ ์œ„ํ—˜์„ ์ถ”์ •ํ•  ์ˆ˜ ์žˆ๋Š” ์ง€ํ‘œ โ€๊ฐ€ ์ ˆ์‹คํžˆ ์š”๊ตฌ๋œ๋‹ค. 2. ํ•ต์‹ฌ ์ด๋ก ์  ๊ธฐ์—ฌ | ๊ตฌ๋ถ„ | ๋‚ด์šฉ | ์˜์˜ | | | | | | ์„ ํ˜• ๊ฐ€์šฐ์‹œ์•ˆ ๋ชจ๋ธ | ๋ ˆ๋ฒ„๋ฆฌ์ง€ ์ ์ˆ˜ (h {ii} x i^top (X

Statistics Machine Learning
No Image

GICDM: Mitigating Hubness for Reliable Distance-Based Generative Model Evaluation

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ ๊ณ ์ฐจ์› ์ž„๋ฒ ๋”ฉ์˜ ๋ณดํŽธ์„ฑ : ์ตœ์‹  ์ด๋ฏธ์ง€ยท์Œ์„ฑยทํ…์ŠคํŠธ ํ‰๊ฐ€์—์„œ๋Š” Inceptionโ€‘V3(2048), DINOv3(4096), CLAP(1024) ๋“ฑ ์ˆ˜์ฒœ ์ฐจ์›์˜ ์‚ฌ์ „ ํ•™์Šต ์ž„๋ฒ ๋”ฉ์„ ์‚ฌ์šฉํ•œ๋‹ค. ํ—ˆ๋ธŒ ํ˜„์ƒ(Hubness) : ๊ณ ์ฐจ์›์—์„œ๋Š” ํŠน์ • ํฌ์ธํŠธ๊ฐ€ ๋‹ค๋ฅธ ํฌ์ธํŠธ๋“ค์˜ kโ€‘nearest neighbor์— ๊ณผ๋„ํ•˜๊ฒŒ ๋“ฑ์žฅํ•œ๋‹ค. ์ด๋Š” ๊ฑฐ๋ฆฌ ๊ธฐ๋ฐ˜ ๋ฉ”ํŠธ๋ฆญ(์˜ˆ: FID, Precisionโ€‘Recall, Coverage)์—์„œ โ€œ๊ทผ์ ‘์„ฑโ€์ด ์˜๋ฏธ๋ฅผ ์žƒ๊ฒŒ ๋งŒ๋“ ๋‹ค. ๊ธฐ์กด ์™„ํ™” ๋ฐฉ๋ฒ•์˜ ํ•œ๊ณ„ : ๋Œ€๋ถ€๋ถ„์˜ ํ—ˆ๋ธŒ ๊ฐ์†Œ ๊ธฐ๋ฒ•์€ inโ€‘sa

Computer Science Machine Learning Model
No Image

GMAIL: Generative Modality Alignment for generated Image Learning

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ ์ƒ์„ฑ ์ด๋ฏธ์ง€์˜ ์ž ์žฌ๋ ฅ : Diffusion, Stableโ€‘Diffusion ๋“ฑ ์ตœ์‹  ์ƒ์„ฑ ๋ชจ๋ธ์€ ๊ณ ํ’ˆ์งˆ ์ด๋ฏธ์ง€๋ฅผ ์ €๋น„์šฉ์œผ๋กœ ๋Œ€๋Ÿ‰ ์ƒ์‚ฐ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ•˜์—ฌ ๋ฐ์ดํ„ฐ ์ฆ๊ฐ•์— ํฐ ๊ธฐ๋Œ€๋ฅผ ๋ชจ์€๋‹ค. ๋ชจ๋‹ฌ๋ฆฌํ‹ฐ ๊ฒฉ์ฐจ : ์ƒ์„ฑ ์ด๋ฏธ์ง€์—๋Š” ์‹œ๊ฐ์  ์•„ํ‹ฐํŒฉํŠธ, ํŽธํ–ฅ, ๋„๋ฉ”์ธโ€‘ํŠน์ • ๋…ธ์ด์ฆˆ๊ฐ€ ์กด์žฌํ•ด, ๋‹จ์ˆœํžˆ ํ”ฝ์…€ ์ˆ˜์ค€์—์„œ ์‹ค์ œ ์ด๋ฏธ์ง€์™€ ๊ต์ฒดํ•˜๋ฉด Genโ€‘Real ๋ชจ๋‹ฌ๋ฆฌํ‹ฐ ์ฐจ์ด ๊ฐ€ ๋ฐœ์ƒํ•œ๋‹ค. ์ด๋Š” ๋ชจ๋ธ์ด ํ•ฉ์„ฑ ๋ฐ์ดํ„ฐ์— ๊ณผ๋„ํžˆ ์ ํ•ฉ(overโ€‘fit)๋˜์–ด ์‹ค์ œ ๋ฐ์ดํ„ฐ์— ๋Œ€ํ•œ ์ผ๋ฐ˜ํ™”๊ฐ€ ์ €ํ•˜๋˜๋Š” ๋ชจ๋“œ ๋ถ•๊ดด ๋ฅผ ์ดˆ๋ž˜ํ•œ๋‹ค. 2. ๊ธฐ์กด ์ ‘๊ทผ๋ฒ•๊ณผ ํ•œ

Computer Science Learning Computer Vision
No Image

Gradient Networks for Universal Magnetic Modeling of Synchronous Machines

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์ „๋™๊ธฐ ์ œ์–ดยท์˜ˆ์ธก ์— ํ•„์ˆ˜์ ์ธ ๊ณ ์ •๋ฐ€ ์ž๊ธฐ ๋ชจ๋ธ์€ ํฌํ™”ยท๊ณต๊ฐ„ ๊ณ ์กฐํŒŒ๋ฅผ ํฌํ•จํ•ด์•ผ ํ•˜์ง€๋งŒ, ์ „ํ†ต์ ์ธ ๋ถ„์„์‹ ์€ ์ฐจ์› ํ™•์žฅ์— ํ•œ๊ณ„๊ฐ€ ์žˆ๊ณ , ๋ฃฉ์—…โ€‘ํ…Œ์ด๋ธ” ์€ ์ฐจ์›์˜ ์ €์ฃผ์™€ ๋น„๋งค๋„๋Ÿฌ์šด ๋ณด๊ฐ„ ๋ฌธ์ œ๋ฅผ ์•ˆ๊ณ  ์žˆ๋‹ค. ๋ธ”๋ž™โ€‘๋ฐ•์Šค NN ์€ ๊ณ ์ฐจ์› ๋งคํ•‘์„ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ•˜์ง€๋งŒ, ๋ฌผ๋ฆฌ์  ์ œ์•ฝ(์—๋„ˆ์ง€ ๋ณด์กด, ๋‹จ์กฐ์„ฑ) ์ด ์—†์œผ๋ฉฐ, ๋Œ€๊ทœ๋ชจ ๋ฐ์ดํ„ฐ์™€ ์™ธ์‚ฝ์— ์ทจ์•ฝํ•˜๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด โ€“ Gradient Network Hamiltonian NN ์€ ์Šค์นผ๋ผ ์—๋„ˆ์ง€ ํ•จ์ˆ˜๋ฅผ ํ•™์Šตํ•˜๊ณ  ์ˆ˜์น˜ ๋ฏธ๋ถ„์œผ๋กœ ๊ธฐ์šธ๊ธฐ๋ฅผ ์–ป์ง€๋งŒ, ๋ฏธ๋ถ„ ์˜ค์ฐจ์™€ ํ•™์Šต ๋ถˆ์•ˆ์ •์„ฑ์„

Network Electrical Engineering and Systems Science Model
No Image

Grammar-based Ordinary Differential Equation Discovery

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ๋™์—ญํ•™ ์‹œ์Šคํ…œ ๋ฐœ๊ฒฌ ์€ ๋ฌผ๋ฆฌยท๊ณตํ•™ ๋ถ„์•ผ์—์„œ ์‹ค์‹œ๊ฐ„ ์ œ์–ด, ๊ณ ์žฅ ์ง„๋‹จ, ์˜ˆ์ธก ์œ ์ง€๋ณด์ˆ˜ ๋“ฑ์— ํ•„์ˆ˜์ ์ด๋‹ค. ๊ธฐ์กด ์ ‘๊ทผ๋ฒ•์€ ์œ ์ „ ํ”„๋กœ๊ทธ๋ž˜๋ฐ(GP) , ์‹ฌ์ธต ํ•™์Šต ๊ธฐ๋ฐ˜ ์‹ฌ๋ณผ๋ฆญ ํšŒ๊ท€(์˜ˆ: ํŠธ๋žœ์Šคํฌ๋จธ) ๋“ฑ์ด๋‹ค. ํ•˜์ง€๋งŒ ๊ณ ์ฐจ์› ํƒ์ƒ‰ ๋น„์šฉ์ด ํฌ๊ณ , ๊ณผ์ ํ•ฉยทํ•ด์„์„ฑ ๋ถ€์กฑ ๋ฌธ์ œ๊ฐ€ ์žˆ๋‹ค. ๋ฌธ๋ฒ•(Grammar) ์„ ํ™œ์šฉํ•˜๋ฉด ๋„๋ฉ”์ธ ์ง€์‹์„ ๊ตฌ์กฐ์ ์œผ๋กœ ์ธ์ฝ”๋”ฉํ•ด ํƒ์ƒ‰ ๊ณต๊ฐ„์„ ์ž์—ฐ์Šค๋Ÿฝ๊ฒŒ ์ œํ•œํ•˜๋ฉด์„œ๋„ ์ถฉ๋ถ„ํžˆ ํ’๋ถ€ํ•˜๊ฒŒ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด ๋ฐ ๋ฐฉ๋ฒ•๋ก  | ๊ตฌ์„ฑ ์š”์†Œ | ์—ญํ•  | ์ฃผ์š” ๊ธฐ๋ฒ• | | | | | | Formal Gram

No Image

Graph neural networks uncover structure and functions underlying the activity of simulated neural assemblies

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ์ฐจ๋ณ„์  | ๊ธฐ์กด ์ ‘๊ทผ | ๋ชฉํ‘œ | ํ•œ๊ณ„ | | | | | | RNN / Transformer ๊ธฐ๋ฐ˜ ์˜ˆ์ธก ๋ชจ๋ธ | ์‹ ๊ฒฝ ํ™œ๋™ ์˜ˆ์ธก | ํ•ด์„ ๊ฐ€๋Šฅ์„ฑ ๋ถ€์กฑ, ๋‰ด๋Ÿฐ ์ด์งˆ์„ฑ ๋ฐ˜์˜ ์–ด๋ ค์›€ | | ๊ธฐ๋Šฅ์  ์—ฐ๊ฒฐ์„ฑ ์ถ”์ • (์˜ˆ: Mi et al., 2021) | ๋ถ€๋ถ„์ ์ธ ์—ฐ๊ฒฐยท์ „์•• ๋ณต์› | ์ „์ฒด ๋ฉ”์ปค๋‹ˆ์ฆ˜(ํ•จ์ˆ˜ยท์™ธ๋ถ€ ์ž…๋ ฅ) ํ†ตํ•ฉ ๋ถˆ๊ฐ€ | | ์„ ํ˜•ยท์‹ฌ๋ณผ๋ฆญ ํšŒ๊ท€ | ํŠน์ • ๋ณ€์ˆ˜ ์ถ”์ • | ๋ณต์žกํ•œ ๋น„์„ ํ˜• ์ƒํ˜ธ์ž‘์šฉ, ๋‹ค์ค‘ ๋‰ด๋Ÿฐ ์œ ํ˜• ๋‹ค๋ฃจ๊ธฐ ์–ด๋ ค์›€ | ๋ณธ ๋…ผ๋ฌธ์€ โ€œ๋™์‹œ ๋ณต์›โ€ ์ด๋ผ๋Š” ์ƒˆ๋กœ์šด ํŒจ๋Ÿฌ๋‹ค์ž„์„ ์ œ์‹œํ•œ๋‹ค. GNN์˜ ๋ฉ”์‹œ์ง€โ€‘ํŒจ์‹ฑ ๊ตฌ์กฐ๋ฅผ ํ™œ์šฉ

Network Quantitative Biology
No Image

Grip as Needed, Glide on Demand: Ultrasonic Lubrication for Robotic Locomotion

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์ „ํ†ต์  ๋งˆ์ฐฐ ์ œ์–ด ํ•œ๊ณ„: ๊ธฐ์กด ๋กœ๋ด‡ ์ด๋™์—์„œ๋Š” ์ •์ƒ๋ ฅ ์กฐ์ ˆ, ๊ตฌ์กฐ์  ์ด๋ฐฉ์„ฑ(๋‚ ๊ฐœ, ๋ธŒ๋Ÿฌ์‹œ ๋“ฑ) ๋˜๋Š” ๊ณ ์ •โ€‘์Šฌ๋ผ์ด๋“œ ๋น„์œจ์„ ์ด์šฉํ•ด ๋งˆ์ฐฐ์„ ์ œ์–ดํ•œ๋‹ค. ์ด๋Ÿฌํ•œ ๋ฐฉ๋ฒ•์€ ํ‘œ๋ฉด ๊ฑฐ์น ๊ธฐยท์Šต์œคยทํƒ„์„ฑ์— ํฌ๊ฒŒ ์˜์กดํ•ด ๋ณตํ•ฉ ํ™˜๊ฒฝ์—์„œ ์„ฑ๋Šฅ์ด ์ €ํ•˜๋œ๋‹ค. ์ดˆ์ŒํŒŒ ์œคํ™œ์˜ ์ฐจ๋ณ„์ : ์ดˆ์ŒํŒŒ ์ง„๋™์œผ๋กœ ์ ‘์ด‰๋ฉด์— ์••์ถ• ์œ ์ฒด๋ง‰์„ ํ˜•์„ฑํ•ด ๋งˆ์ฐฐ ๊ณ„์ˆ˜๋ฅผ ์ง์ ‘ ์กฐ์ ˆํ•œ๋‹ค. ๋งˆ์ฐฐ์„ โ€˜๊ทธ๋ฆฝ(๊ณ ๋งˆ์ฐฐ)โ€™ โ†” โ€˜์Šฌ๋ฆฝ(์ €๋งˆ์ฐฐ)โ€™ ์ƒํƒœ๋กœ ๋น ๋ฅด๊ฒŒ ์ „ํ™˜ํ•  ์ˆ˜ ์žˆ์–ด, ๋ฌผ๋ฆฌ์  ๊ตฌ์กฐ ๋ณ€ํ™” ์—†์ด๋„ ์‹ค์‹œ๊ฐ„ ์ œ์–ด๊ฐ€ ๊ฐ€๋Šฅํ•˜๋‹ค. 2. ํ•ต์‹ฌ ๊ธฐ์ˆ ยท์„ค๊ณ„ | ๋ชจ๋“ˆ | ํ˜•ํƒœ | ์ฃผ์š” ํŒŒ๋ผ๋ฏธ

Robotics Computer Science
No Image

Group Interventions on Deep Networks for Causal Discovery in Subsystems

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์ธ๊ณผ ํƒ์ƒ‰์˜ ํ•œ๊ณ„ : ์ „ํ†ต์ ์ธ ์ธ๊ณผ ํƒ์ƒ‰ ๊ธฐ๋ฒ•(Granger, PCMCI, NOTEARS ๋“ฑ)์€ ์ฃผ๋กœ ๋ณ€์ˆ˜ ์Œ ๊ฐ„ ์ธ๊ณผ์„ฑ์„ ํƒ์ƒ‰ํ•œ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์‹ค์ œ ์‹œ์Šคํ…œ(๋‡Œ, ๊ธฐํ›„ ๋“ฑ)์€ ๋‹ค์ˆ˜์˜ ๋ณ€์ˆ˜๋“ค์ด ์„œ๋กœ ์–ฝํžŒ ์„œ๋ธŒ์‹œ์Šคํ…œ ํ˜•ํƒœ๋กœ ์ž‘๋™ํ•œ๋‹ค. ๊ทธ๋ฃน ์ˆ˜์ค€ ์ธ๊ณผ์„ฑ : ๊ทธ๋ฃน(์„œ๋ธŒ์‹œ์Šคํ…œ) ๊ฐ„ ์ธ๊ณผ ๊ด€๊ณ„๋ฅผ ํŒŒ์•…ํ•˜๋ฉด, ๊ฐœ๋ณ„ ๋ณ€์ˆ˜ ์ˆ˜์ค€์—์„œ ๋†“์น  ์ˆ˜ ์žˆ๋Š” ์ง‘ํ•ฉ์  ํšจ๊ณผ ์™€ ์ „๋‹ฌ ๋ฉ”์ปค๋‹ˆ์ฆ˜ ์„ ํฌ์ฐฉํ•  ์ˆ˜ ์žˆ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด ๋ฐ ๋ฐฉ๋ฒ•๋ก  | ๋‹จ๊ณ„ | ๋‚ด์šฉ | ์ฃผ์š” ๊ธฐ๋ฒ• | | | | | | 1. ๊ตฌ์กฐ์  ๋”ฅ ๋ชจ๋ธ๋ง | ๋ชจ๋“  ์‹œ๊ณ„์—ด์„ ๊ทธ๋ฃนํ™”

System Network
No Image

HAL-MLE Log-Splines Density Estimation (Part I: Univariate)

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์ „ํ†ต์  ์ปค๋„ ๋ฐ€๋„ ์ถ”์ •(KDE) ์€ ์ฐจ์› ์ €์ฃผ์™€ ๊ธ‰๊ฒฉํžˆ ๋ณ€ํ•˜๋Š” ๊ตฌ๊ฐ„์—์„œ์˜ ๊ณผโ€‘/๊ณผ์†Œ ํ‰ํ™œํ™” ๋ฌธ์ œ๋ฅผ ์•ˆ๊ณ  ์žˆ๋‹ค. TVโ€‘ํŒจ๋„ํ‹ฐ ๋Š” ํšŒ๊ท€ยท์Šคํ”Œ๋ผ์ธ ๋ถ„์•ผ์—์„œ ์ด๋ฏธ ์„ฑ๊ณต์ ์œผ๋กœ ์ ์šฉ๋ผ, ๊ธ‰๊ฒฉํ•œ ๋ณ€ํ™”๋ฅผ ์–ต์ œํ•˜๋ฉด์„œ๋„ ์ ์‘์  ์ž์œ ๋„๋ฅผ ์ œ๊ณตํ•œ๋‹ค. ๊ธฐ์กด logโ€‘spline (Kooperberg & Stone, 1992) ์€ ์–‘์„ฑ ๋ณด์žฅ์„ ์œ„ํ•ด ์ง€์ˆ˜ ๋งํฌ๋ฅผ ์‚ฌ์šฉํ•˜์ง€๋งŒ, ๋‹ค๋ณ€๋Ÿ‰ ํ™•์žฅ์€ ๊ท ์ผ ๋…ธ๋“œ ๊ฐ€์ •๊ณผ ๊ณ ์ฐจ ์—ฐ์†์„ฑ ์š”๊ตฌ ๋•Œ๋ฌธ์— ์–ด๋ ค์›€์„ ๊ฒช๋Š”๋‹ค. 2. HALโ€‘MLE์™€ ๊ธฐ์กด ๋ฐฉ๋ฒ•์˜ ์—ฐ๊ฒฐ ๊ณ ๋ฆฌ | ๋ฐฉ๋ฒ• | ๊ธฐ๋ณธ ์•„์ด๋””์–ด | ํ•จ์ˆ˜ ํด๋ž˜์Šค

Mathematics
No Image

Hardware-accelerated graph neural networks: an alternative approach for neuromorphic event-based audio classification and keyword spotting on SoC FPGA

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์—ฃ์ง€ AI์™€ ์ „๋ ฅ ์ œ์•ฝ : IoTยท์›จ์–ด๋Ÿฌ๋ธ” ๋“ฑ ๋ฐฐํ„ฐ๋ฆฌ ๊ตฌ๋™ ์žฅ์น˜์—์„œ ์‹ค์‹œ๊ฐ„ ์˜ค๋””์˜ค ์ธ์‹์€ ์ „๋ ฅยท์ง€์—ฐ ์ œํ•œ์ด ํฌ๋‹ค. ๊ธฐ์กด GPUยทCPU ๊ธฐ๋ฐ˜ ์†”๋ฃจ์…˜์€ ์ „๋ ฅ ์†Œ๋ชจ๊ฐ€ ํฌ๊ณ , ์ง€์—ฐ์ด ํ—ˆ์šฉ ๋ฒ”์œ„๋ฅผ ์ดˆ๊ณผํ•œ๋‹ค. ์ด๋ฒคํŠธโ€‘๊ธฐ๋ฐ˜ ์„ผ์„œ : ์ธ๊ณต ์™€์šฐ๋Š” ์‹ ํ˜ธ ๋ณ€ํ™”๊ฐ€ ์žˆ์„ ๋•Œ๋งŒ ์ด๋ฒคํŠธ๋ฅผ ๋ฐœ์ƒ์‹œ์ผœ ๋ฐ์ดํ„ฐ ํฌ์†Œ์„ฑ์„ ํฌ๊ฒŒ ๋†’์ธ๋‹ค. ์ด๋Š” ๋ฉ”๋ชจ๋ฆฌยท์—ฐ์‚ฐ๋Ÿ‰ ๊ฐ์†Œ์™€ ์ „๋ ฅ ์ ˆ๊ฐ์œผ๋กœ ์ง๊ฒฐ๋œ๋‹ค. ๊ทธ๋ž˜ํ”„ ์‹ ๊ฒฝ๋ง์˜ ์žฅ์  : ๊ธฐ์กด ์ŠคํŒŒ์ดํ‚น ์‹ ๊ฒฝ๋ง(SNN)์€ ๋น„๊ฒฐ์ •์  ๋ฉ”๋ชจ๋ฆฌ ์ ‘๊ทผ ํŒจํ„ด์œผ๋กœ ํ•˜๋“œ์›จ์–ด ์ตœ์ ํ™”๊ฐ€ ์–ด๋ ค์› ๋‹ค. ์ด๋ฒคํŠธโ€‘๊ทธ๋ž˜ํ”„ ์‹ ๊ฒฝ๋ง์€ ๊ทธ๋ž˜ํ”„ ๊ตฌ์กฐ๊ฐ€

Computer Science Network Machine Learning
No Image

Harnessing Implicit Cooperation: A Multi-Agent Reinforcement Learning Approach Towards Decentralized Local Energy Markets

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ ์—๋„ˆ์ง€ ์ „ํ™˜ 3๋Œ€ ์ถ•(ํƒˆํƒ„์†Œยท๋””์ง€ํ„ธยท๋ถ„์‚ฐ) ์œผ๋กœ DER(๋ถ„์‚ฐํ˜• ์—๋„ˆ์ง€ ์ž์›)์˜ ๊ธ‰์ฆ โ†’ ์ „ํ†ต์  ์ค‘์•™์ง‘์ค‘์‹ ์ œ์–ด๋Š” ๊ณ„์‚ฐ ๋ณต์žก๋„ ํญ๋ฐœ ๋ฐ ๋‹จ์ผ ์žฅ์• ์ (singleโ€‘pointโ€‘ofโ€‘failure) ๋ฌธ์ œ์— ์ง๋ฉด. ๋กœ์ปฌ ์—๋„ˆ์ง€ ์‹œ์žฅ(LEM)์€ ์ด๋Ÿฌํ•œ ๋ณต์žก์„ฑ์„ ์™„ํ™”ํ•˜๊ธฐ ์œ„ํ•œ ์šด์˜ ํ”„๋ ˆ์ž„์›Œํฌ์ด์ง€๋งŒ, ํ™•์žฅ์„ฑ, ํ”„๋ผ์ด๋ฒ„์‹œ, ๊ณต๊ธ‰โ€‘์ˆ˜์š” ๊ท ํ˜• ์„ ๋™์‹œ์— ๋งŒ์กฑ์‹œํ‚ค๋Š” ์†”๋ฃจ์…˜์ด ๋ถ€์žฌ. ๊ธฐ์กด MARL ์ ‘๊ทผ๋ฒ•์€ ์ค‘์•™์ง‘์ค‘์‹ ํ•™์Šต(CTCE) ์— ์˜์กดํ•ด ์ „์—ญ ์ƒํƒœ์™€ ๊ฐœ์ธ ๋น„์šฉ ํ•จ์ˆ˜๋ฅผ ๊ณต์œ ํ•จ์œผ๋กœ์จ ํ”„๋ผ์ด๋ฒ„์‹œ๋ฅผ ์นจํ•ดํ•˜๊ฑฐ๋‚˜, ์™„์ „ ๋ถ„์‚ฐ ํ•™์Šต

Learning Electrical Engineering and Systems Science
No Image

Heat Equation driven by mixed local-nonlocal operators with non-regular space-dependent coefficients

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ํ˜ผํ•ฉ ์—ฐ์‚ฐ์ž (L 0 Delta+( Delta)^s) ์€ ์ตœ๊ทผ ํ™•์‚ฐยท์ „์ด ํ˜„์ƒ์˜ ๋ณตํ•ฉ ๋ฉ”์ปค๋‹ˆ์ฆ˜์„ ๋ชจ๋ธ๋งํ•˜๋Š” ๋ฐ ๋„๋ฆฌ ์‚ฌ์šฉ๋œ๋‹ค. ํŠนํžˆ, ์ƒํƒœํ•™ยท์ƒ๋ฌผํ•™ ์—์„œ ์งง์€ ๊ฑฐ๋ฆฌ์˜ ๋ธŒ๋ผ์šด ์šด๋™๊ณผ ๊ธด ๊ฑฐ๋ฆฌ์˜ Lรฉvy ์ ํ”„๋ฅผ ๋™์‹œ์— ๊ณ ๋ คํ•˜๋Š” ๊ฒƒ์ด ํ˜„์‹ค์ ์ธ ์ด๋™ ํŒจํ„ด์„ ์„ค๋ช…ํ•œ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ๋Š” ์ฃผ๋กœ ๊ท ์ผํ•˜๊ฑฐ๋‚˜ ์ถฉ๋ถ„ํžˆ ์ •์น™ํ•œ ๊ณ„์ˆ˜ (์˜ˆ: ์ƒ์ˆ˜ (a,b,c))์— ํ•œ์ •๋ผ ์žˆ์—ˆ์œผ๋ฉฐ, ๋ถˆ๊ทœ์น™ ๋งค์งˆ (์˜ˆ: ๊ธ‰๊ฒฉํ•œ ๊ฒฝ๊ณ„, ์ ๊ทผ์  ํก์ˆ˜, ํ˜น์€ ์ธก์ • ๋ถˆ๊ฐ€๋Šฅํ•œ ๋ฌผ์„ฑ)์—์„œ๋Š” ํ•ด์„์  ๋„๊ตฌ๊ฐ€ ๋ถ€์กฑํ–ˆ๋‹ค. 2. ์ฃผ์š” ๊ธฐ์—ฌ | ๊ตฌ๋ถ„ | ๋‚ด์šฉ |

Mathematics
No Image

Hensel minimality, $p$-adic exponentiation and Tate uniformization

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ oโ€‘minimal์„ฑ ์€ ์‹คยท๋ณต์†Œ ํ•ด์„์—์„œ โ€œ๊ฑฐ์นœโ€ ํ˜„์ƒ์„ ์–ต์ œํ•˜๊ณ  ์ •๋ฐ€ํ•œ ๊ธฐํ•˜ยท์ˆ˜๋ก ์  ์ถ”๋ก ์„ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ•œ๋‹ค. ๋น„์•„ํ‚ค๋ฉ”๋ฐ์•ˆ(ํŠนํžˆ $p$โ€‘adic) ์„ธ๊ณ„์—์„œ๋„ ๋น„์Šทํ•œ tameness ๊ฐ€ ํ•„์š”ํ–ˆ์œผ๋ฉฐ, ๊ธฐ์กด์˜ Pโ€‘minimal , Cโ€‘minimal ์ด๋ก ์€ ์ œํ•œ์ ์ธ ์ ์šฉ ๋ฒ”์œ„์— ๋จธ๋ฌผ๋ €๋‹ค. ์ตœ๊ทผ ์ œ์•ˆ๋œ 1โ€‘hโ€‘minimality (Cluckersโ€‘Halupczokโ€‘Ribeiroโ€‘Kuhlmann ๋“ฑ)๋Š” ๋ถ„๋ฆฌ๋œ Weierstrass ์‹œ์Šคํ…œ ์„ ํฌํ•จํ•˜๋Š” ๋„“์€ ํด๋ž˜์Šค์— ์ ์šฉ ๊ฐ€๋Šฅํ•˜๊ณ , Jacobian property , ์•Œ์ œ

Mathematics
No Image

Hermes: Large DEL Datasets Train Generalizable Protein-Ligand Binding Prediction Models

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๋ฐ์ดํ„ฐ ํŽธํ–ฅ ๋ฌธ์ œ : BindingDB, ChEMBL ๋“ฑ ๊ณต๊ฐœ ์นœํ™”๋„ ๋ฐ์ดํ„ฐ๋Š” ์‹คํ—˜์‹คยท์–ด์„ธ์ดยทํ”„๋กœํ† ์ฝœ์ด ๋‹ค์–‘ํ•ด ํ‘œ์ค€ํ™”๊ฐ€ ์–ด๋ ค์šฐ๋ฉฐ, ๋ชจ๋ธ์ด ํ•™์Šต ์‹œ ํŽธํ–ฅ์„ ๊ทธ๋Œ€๋กœ ํ•™์Šตํ•˜๊ฒŒ ๋œ๋‹ค. DEL์˜ ์žฅ์  : ๋™์ผ ํ”„๋กœํ† ์ฝœ๋กœ ์ˆ˜์‹ญ์–ต ํ™”ํ•ฉ๋ฌผ์„ ๋™์‹œ์— ์Šคํฌ๋ฆฌ๋‹ํ•˜๋ฏ€๋กœ, ์ผ๊ด€๋œ ๋ผ๋ฒจ(Enrichment) ๊ณผ ๊ด‘๋ฒ”์œ„ํ•œ ํ™”ํ•™ยท๋‹จ๋ฐฑ์งˆ ๊ณต๊ฐ„ ์„ ๋™์‹œ์— ์ œ๊ณตํ•œ๋‹ค. 2. ์ฃผ์š” ๊ธฐ์—ฌ | ๊ตฌ๋ถ„ | ๋‚ด์šฉ | | | | | ๋ฐ์ดํ„ฐ | 239๊ฐœ ๋‹จ๋ฐฑ์งˆ(โ‰ˆ2/3์ด ํ‚ค๋‚˜์ œ)ยท6.5M ํ™”ํ•ฉ๋ฌผ(Kin0) DEL ์Šคํฌ๋ฆฌ๋‹ ๋ฐ์ดํ„ฐ โ†’ ํ˜„์žฌ ๊ณต๊ฐœ๋œ DEL ์ค‘

Data Quantitative Biology Model
No Image

Heterogeneous Robot Collaboration in Unstructured Environments with Grounded Generative Intelligence

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ ์ด์งˆ ๋กœ๋ด‡ ํŒ€ : ์„œ๋กœ ๋‹ค๋ฅธ ์ด๋™ยท์„ผ์„œยท์ž‘์—… ๋Šฅ๋ ฅ์„ ๊ฐ€์ง„ ๋กœ๋ด‡๋“ค์ด ํ˜‘์—…ํ•ด์•ผ ํ•˜๋Š” ์ƒํ™ฉ์€ ์žฌ๋‚œ ๊ตฌ์กฐ, ํƒ์‚ฌ, ๋ฌผ๋ฅ˜ ๋“ฑ์—์„œ ๋นˆ๋ฒˆํžˆ ๋ฐœ์ƒํ•œ๋‹ค. ๋ฌด๊ตฌ์กฐ ํ™˜๊ฒฝ : ์‚ฌ์ „ ์ง€๋„ยท์ •ํ™•ํ•œ ๋ชจ๋ธ์ด ์—†๋Š”, ๋™์ ์œผ๋กœ ๋ณ€ํ•˜๋Š” ํ™˜๊ฒฝ์—์„œ๋Š” ์ „ํ†ต์ ์ธ ๊ณ„ํšโ€‘์‹คํ–‰ ํŒŒ์ดํ”„๋ผ์ธ์ด ์‰ฝ๊ฒŒ ๋ถ•๊ดดํ•œ๋‹ค. ์ƒ์„ฑํ˜• ์ธํ…”๋ฆฌ์ „์Šค : LLM์„ ์ด์šฉํ•ด ์ž์—ฐ์–ด๋กœ ๋œ ๋ฏธ์…˜์„ ์ž๋™์œผ๋กœ ํ•ด์„ยท๋ถ„ํ•ดํ•˜๋Š” ์‹œ๋„๊ฐ€ ๋Š˜๊ณ  ์žˆ์ง€๋งŒ, ๋Œ€๋ถ€๋ถ„ ๊ตฌ์กฐํ™”๋œ ํ™˜๊ฒฝ(์˜ˆ: ์‹ค๋‚ด ์ง€๋„, ์ •ํ•ด์ง„ ์ž‘์—… ๊ณต๊ฐ„)์—์„œ๋งŒ ๊ฒ€์ฆ๋ผ ์™”๋‹ค. 2. ํ•ต์‹ฌ ๊ธฐ์—ฌ | ๊ตฌ๋ถ„ | ๊ธฐ์กด ์—ฐ๊ตฌ์™€ ์ฐจ๋ณ„์  | ๊ตฌ์ฒด์  ๊ธฐ์—ฌ

No Image

Hidden Markov Individual-level Models of Infectious Disease Transmission

| ํ•ญ๋ชฉ | ๋‚ด์šฉ ๋ฐ ํ‰๊ฐ€ | | | | | ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ | ๊ฐœ์ธโ€‘์ˆ˜์ค€ ์ „์—ผ๋ณ‘ ๋ชจ๋ธ์€ ๊ฐ์ˆ˜์„ฑยท์ „์—ผ์„ฑ ์ด์งˆ์„ฑ์„ ๋ฐ˜์˜ํ•  ์ˆ˜ ์žˆ์–ด ์—ญํ•™ยท๊ณต์ค‘๋ณด๊ฑด์— ํ•„์ˆ˜์ ์ด์ง€๋งŒ, ๊ด€์ธก ๋ฐ์ดํ„ฐ๊ฐ€ โ€œ๊ฐ์ง€ ์‹œ์ โ€ ํ•˜๋‚˜๋งŒ ์ œ๊ณต๋˜๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋งŽ๋‹ค.<br> ๊ธฐ์กด DAโ€‘๊ธฐ๋ฒ•์€ ๋ณต์žกํ•œ reversibleโ€‘jump MCMC๊ฐ€ ํ•„์š”ํ•˜๊ฑฐ๋‚˜, ๊ฐ์ง€ ๊ฐ์—ผยท์ œ๊ฑฐ๋ผ๋Š” ๋น„ํ˜„์‹ค์  ๊ฐ€์ •์„ ๊ฐ•์š”ํ•œ๋‹ค. | | ํ•ต์‹ฌ ์•„์ด๋””์–ด | ๊ฒฐํ•ฉ ์ˆจ์€ ๋งˆ์ฝ”ํ”„ ๋ชจ๋ธ (coupled HMM)์„ ์‚ฌ์šฉํ•ด ๊ฐ ๊ฐœ์ธ์˜ ๊ฐ์—ผยท์ œ๊ฑฐ ์ƒํƒœ๋ฅผ ์ˆจ์€ ์ƒํƒœ๋กœ ๋‘๊ณ , ๊ด€์ธก ๋ชจ๋ธ ์„ โ€œ์ฒซ ์–‘์„ฑ ์ „๊นŒ์ง€๋Š” ๊ฒ€์ถœ ํ™•๋ฅ  ฮธ, ์ด

Statistics Applications Model
No Image

Hierarchical parameter estimation for distributed networked systems: a dynamic consensus approach

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์ค‘์•™์ง‘์ค‘์‹ vs. ๋ถ„์‚ฐ์‹ : ์ค‘์•™์ง‘์ค‘์‹ ์ถ”์ •์€ ๋Œ€์—ญํญยท์—ฐ์‚ฐ๋Ÿ‰ยท๋‚ด๊ฒฐํ•จ์„ฑ ์ธก๋ฉด์—์„œ ํ•œ๊ณ„๊ฐ€ ์žˆ๋‹ค. ํŠนํžˆ ์—์ด์ „ํŠธ ์ˆ˜๊ฐ€ ๋Š˜์–ด๋‚ ์ˆ˜๋ก ํ†ต์‹  ๋ถ€ํ•˜์™€ ๋‹จ์ผ ์žฅ์• ์ (single point of failure) ๋ฌธ์ œ๊ฐ€ ์‹ฌํ™”๋œ๋‹ค. ๊ธฐ์กด ๋ถ„์‚ฐ ์ ‘๊ทผ๋ฒ•์˜ ํ•œ๊ณ„ : โ€˜consensus + innovationsโ€™ ๋ฐฉ์‹์€ ํ•ฉ์˜์™€ ์ถ”์ •์ด ๊ธด๋ฐ€ํžˆ ๊ฒฐํ•ฉ๋ผ Lyapunov ๋ถ„์„์ด ๋ณต์žกํ•˜๊ณ , ์ƒˆ๋กœ์šด ์ถ”์ •๊ธฐ๋ฅผ ๋„์ž…ํ•˜๊ฑฐ๋‚˜ ์–‘์žํ™”ยทํ† ํด๋กœ์ง€ ๋ณ€ํ™”๋ฅผ ๊ณ ๋ คํ•  ๋•Œ ์ „์ฒด ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์žฌ์„ค๊ณ„ํ•ด์•ผ ํ•˜๋Š” ๋น„ํšจ์œจ์„ฑ์ด ์žˆ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด โ€“ ๊ณ„์ธตํ˜• ๊ตฌ์กฐ ํ•ฉ์˜์™€

System Network Electrical Engineering and Systems Science
No Image

Highly Localised Droplet Clustering in Shallow Cumulus Clouds

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๊ตฌ๋ฆ„ ๋ฏธ์„ธ๋ฌผ๋ฆฌํ•™์˜ ํ•ต์‹ฌ ๋ณ€์ˆ˜ : ๋ฌผ๋ฐฉ์šธ์˜ ์„ฑ์žฅยท์†Œ๋ฉธยท๋ถ„ํฌ๋Š” ๊ตฌ๋ฆ„์˜ ๋ฐ˜์‚ฌยทํก์ˆ˜ ํŠน์„ฑ๊ณผ ์ง์ ‘ ์—ฐ๊ฒฐ๋ผ ๊ธฐํ›„ ๋ชจ๋ธ์˜ ๋ถˆํ™•์‹ค์„ฑ์„ ํฌ๊ฒŒ ์ขŒ์šฐํ•œ๋‹ค. ๊ธฐ์กด ์ธก์ • ํ•œ๊ณ„ : ํ•ด์–‘ ์ธต์šด์—์„œ์˜ ๊ธฐ์กด ์—ฐ๊ตฌ๋Š” ์ˆ˜ ํ‚ฌ๋กœ๋ฏธํ„ฐ ๊ทœ๋ชจ์˜ ํ‰๊ท ํ™”๋œ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ–ˆ์œผ๋ฉฐ, ์ด๋Š” ๋ฏธ์„ธํ•œ ๊ณต๊ฐ„์  ๋ณ€๋™์„ฑ์„ ํฌ์„์‹œํ‚ฌ ์œ„ํ—˜์ด ์žˆ๋‹ค. ์ƒˆ๋กœ์šด ์ ‘๊ทผ๋ฒ• : โ€œCloudKiteโ€์ด๋ผ๋Š” ๊ณ ๊ณ ๋„ ๋ฌด์ธ๋น„ํ–‰์ฒด์— ํƒ‘์žฌ๋œ ์ดˆ๊ณ ํ•ด์ƒ๋„ ํ™€๋กœ๊ทธ๋ž˜ํ”ฝ ์นด๋ฉ”๋ผ๋ฅผ ์ด์šฉํ•ด, 12 cm ๊ฐ„๊ฒฉ์œผ๋กœ ๋…๋ฆฝ์ ์ธ ์ธก์ • ๋ถ€ํ”ผ๋ฅผ ํ™•๋ณดํ•จ์œผ๋กœ์จ ๋ฐ€๋ฆฌ๋ฏธํ„ฐ ์ดํ•˜ ์˜ ์Šค์ผ€์ผ์„ ์ง์ ‘ ํƒ์ƒ‰ํ–ˆ๋‹ค. 2. ์‹คํ—˜ ์„ค๊ณ„

No Image

Household size can explain 40% of the variance in cumulative COVID-19 incidence across Europe

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ชฉ์  ํ•ต์‹ฌ ์งˆ๋ฌธ : ์œ ๋Ÿฝ ๊ฐ๊ตญ์˜ COVIDโ€‘19 ๋ฐœ์ƒ๋ฅ  ์ฐจ์ด๋ฅผ ์„ค๋ช…ํ•˜๋Š” ๋ฐ ๊ฐ€๊ตฌ ๊ทœ๋ชจ๋งŒ์œผ๋กœ ์–ผ๋งˆ๋งŒํผ ์„ค๋ช…ํ•  ์ˆ˜ ์žˆ๋Š”๊ฐ€? ๋™๊ธฐ : ๊ธฐ์กด NPI๋Š” ์ฃผ๋กœ ๊ฐ€๊ตฌ ์™ธ ์ „ํŒŒ๋ฅผ ์ฐจ๋‹จํ•˜์ง€๋งŒ, ๊ฐ€๊ตฌ ๋‚ด ์ „ํŒŒ๋Š” ์ง€์†์ ยท๊ณ ์œ„ํ—˜ ์ ‘์ด‰์œผ๋กœ ์ธํ•ด ์–ต์ œ ํšจ๊ณผ๊ฐ€ ์ œํ•œ์ ์ด๋‹ค. ๊ฐ€๊ตฌ ๊ตฌ์กฐ๊ฐ€ ๋‹ค๋ฅธ ๊ตญ๊ฐ€ ๊ฐ„ ์ „์—ผ๋ณ‘ ์—ญํ•™ ์ฐจ์ด๋ฅผ ์„ค๋ช…ํ•  ์ˆ˜ ์žˆ๋Š”์ง€ ๊ฒ€์ฆํ•˜๊ณ ์ž ํ•จ. 2. ๋ฐฉ๋ฒ•๋ก  | ๋‹จ๊ณ„ | ๋‚ด์šฉ | ํ•ต์‹ฌ ํฌ์ธํŠธ | | | | | | ์ˆ˜ํ•™์  ํ”„๋ ˆ์ž„์›Œํฌ | ๊ฐ€๊ตฌ ๋‚ด ์ „ํŒŒ์™€ ๊ฐ€๊ตฌ ์™ธ ์ „ํŒŒ๋ฅผ ๋ถ„๋ฆฌ. ๊ฐ€๊ตฌ ์žฌ์ƒ์‚ฐ์ˆ˜ R H R out ร— B ๋กœ ์ •์˜. | B๋Š” ๊ฐ€

Quantitative Biology
No Image

How Much Does Machine Identity Matter in Anomalous Sound Detection at Test Time?

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ ASD์™€ DCASE ์ฑŒ๋ฆฐ์ง€ : ์ตœ๊ทผ DCASE ๋Œ€ํšŒ๋Š” ๋„๋ฉ”์ธ ๋ณ€๋™์— ๊ฐ•์ธํ•œ ASD ๋ชจ๋ธ์„ ์š”๊ตฌํ•˜๋ฉฐ, ๋งŽ์€ ์—ฐ๊ตฌ๊ฐ€ ๋„๋ฉ”์ธ ์ ์‘ยท์ผ๋ฐ˜ํ™”์— ์ง‘์ค‘ํ•˜๊ณ  ์žˆ๋‹ค. ์ˆจ์€ ์ „์ œ : ํ˜„์žฌ ๋Œ€๋ถ€๋ถ„์˜ ํ‰๊ฐ€ ํ”„๋กœํ† ์ฝœ์€ ํ…Œ์ŠคํŠธ ์‹œ์ ์— ๊ฐ ๋…น์Œ์ด ์–ด๋А ๊ธฐ๊ณ„์— ์†ํ•˜๋Š”์ง€ ์•Œ ์ˆ˜ ์žˆ๋‹ค๊ณ  ๊ฐ€์ •ํ•œ๋‹ค(โ€˜machineโ€‘wiseโ€™ ํ‰๊ฐ€). ์‹ค์ œ ํ˜„์žฅ์˜ ๊ดด๋ฆฌ : ๊ณต์žฅ์—์„œ๋Š” ๋™์ผ ๋ชจ๋ธ์˜ ๊ธฐ๊ณ„๊ฐ€ ์—ฌ๋Ÿฌ ๋Œ€ ๋ฐฐ์น˜๋˜๊ณ , ๋งˆ์ดํฌ ํ•˜๋‚˜๋งŒ์œผ๋กœ๋Š” ์†Œ๋ฆฌ ์ถœ์ฒ˜๋ฅผ ๊ตฌ๋ถ„ํ•˜๊ธฐ ์–ด๋ ต๋‹ค. ์ „์šฉ ์„ผ์„œยท๋งˆ์ดํฌ ์–ด๋ ˆ์ด๋ฅผ ์„ค์น˜ํ•˜๋ฉด ๋น„์šฉยท๋ณต์žก๋„๊ฐ€ ๊ธ‰์ฆํ•œ๋‹ค. 2. ์ œ์•ˆ๋œ ํ‰๊ฐ€ ํ”„๋กœํ† 

Audio Processing Electrical Engineering and Systems Science Detection
No Image

How negative feedback from filamentous actin affects cell shapes and motility

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ์งˆ๋ฌธ ํ•ต์‹ฌ ๋ฌธ์ œ : Fโ€‘์•กํ‹ด์ด GTPase ํ™œ์„ฑํ™”๋ฅผ ์–ต์ œํ•˜๋Š” ์Œ์„ฑ ํ”ผ๋“œ๋ฐฑ์ด ์„ธํฌ์˜ ์ด๋™ยทํ˜•ํƒœ ์ „ํ™˜์— ์–ด๋–ค ๋ฉ”์ปค๋‹ˆ์ฆ˜์œผ๋กœ ์ž‘์šฉํ•˜๋Š”๊ฐ€? ์‹คํ—˜์  ๋™๊ธฐ : ๋‹ค์–‘ํ•œ ์„ธํฌ(ํ˜ธ์ค‘๊ตฌ, ์ผ€๋ผํ‹ฐ์‚ฌ์ดํŠธ, ๋””ํ†ก์‹œ์›€ ๋“ฑ)์—์„œ ๊ด€์ฐฐ๋˜๋Š” โ€œ์•กํ‹ด ํŒŒ๋™โ€(์ง„๋™, ์ „ํŒŒ, ํŒŒ๋™โ€‘๋ฆฌํ•‘ ๋“ฑ) ํ˜„์ƒ์„ ์ด๋ก ์ ์œผ๋กœ ํ†ตํ•ฉํ•˜๊ณ , ํŒŒ๋™์ด ์„ธํฌ ์ˆ˜์ค€์—์„œ ์–ด๋–ป๊ฒŒ ํ˜•ํƒœ์™€ ์šด๋™์„ ์กฐ์ ˆํ•˜๋Š”์ง€ ๊ทœ๋ช…ํ•˜๊ณ ์ž ํ•จ. 2. ๋ชจ๋ธ ๊ตฌ์„ฑ | ๊ตฌ์„ฑ ์š”์†Œ | ๋ณ€์ˆ˜ | ์ฃผ์š” ๊ฐ€์ • | | | | | | GTPase ํ™œ์„ฑํ˜• | (u(x,t)) | ๋ง‰์— ๊ฒฐํ•ฉ, ํ™•์‚ฐ ๊ณ„์ˆ˜ (D u) ์ž‘์Œ

Quantitative Biology
No Image

How Reliable is Your Service at the Extreme Edge? Analytical Modeling of Computational Reliability

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ XEC ๋Š” ๊ธฐ์กด ์—ฃ์ง€ ์„œ๋ฒ„๊ฐ€ ๊ฐ๋‹นํ•˜๊ธฐ ์–ด๋ ค์šด ์ดˆ์ €์ง€์—ฐยท๊ณ ์ฒ˜๋ฆฌ๋Ÿ‰ AI ์ŠคํŠธ๋ฆฌ๋ฐ ์„œ๋น„์Šค๋ฅผ ์†Œ๋น„์ž ๋””๋ฐ”์ด์Šค ํ’€์— ๋ถ„์‚ฐ์‹œํ‚จ๋‹ค. ๊ธฐ์กด ์‹ ๋ขฐ์„ฑ ์—ฐ๊ตฌ๋Š” ์ด์ง„ ์žฅ์• (ON/OFF) ๋ชจ๋ธ์— ์ดˆ์ ์„ ๋งž์ถ”์—ˆ์œผ๋ฉฐ, ์—ฐ์‚ฐ๋Ÿ‰ ๋ณ€๋™ ์„ ๊ณ ๋ คํ•˜์ง€ ์•Š๋Š”๋‹ค. ๋”ฐ๋ผ์„œ โ€œ๋””๋ฐ”์ด์Šค๊ฐ€ ์ผ์ • ์‹œ๊ฐ„ ๋™์•ˆ ์š”๊ตฌ๋œ ์—ฐ์‚ฐ๋Ÿ‰์„ ์ง€์†์ ์œผ๋กœ ์ œ๊ณตํ•  ํ™•๋ฅ โ€์„ ์ •์˜ยท๋ถ„์„ํ•˜๋Š” ๊ฒƒ์ด ํ•„์ˆ˜์ ์ด๋‹ค. 2. ํ•ต์‹ฌ ๊ธฐ์—ฌ | ๋ฒˆํ˜ธ | ๋‚ด์šฉ | | | | | โ‘  | ๋‹จ์ผ ๋””๋ฐ”์ด์Šค์— ๋Œ€ํ•œ ์ปดํ“จํŒ… ์‹ ๋ขฐ์„ฑ ์„ MI์™€ ํžˆ์Šคํ† ๋ฆฌ์ปฌ ๋‘ ์ •๋ณด ์ฒด๊ณ„์—์„œ ํ์‡„ํ˜• ์‹์œผ๋กœ ๋„์ถœ. | | โ‘ก | ๋‹ค์ค‘

Model Computer Science Distributed Computing
No Image

How to Train a Shallow Ensemble

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ MLIP์˜ ๋ถˆํ™•์‹ค์„ฑ : ๋จธ์‹ ๋Ÿฌ๋‹ ๋ชจ๋ธ์€ ํ†ต๊ณ„์  ํ•™์Šต ํ”„๋ ˆ์ž„์›Œํฌ์— ๊ธฐ๋ฐ˜ํ•˜๋ฏ€๋กœ ์—ํ”ผ์Šคํ…Œ๋ฏน(์ง€์‹ ๋ถ€์กฑ)ยท์•Œ๋ ˆ์•„ํŠธ๋ฆฌ(๋ฐ์ดํ„ฐ ์žก์Œ)ยท๋ชจ๋ธ ๋ฏธ์Šค์ŠคํŽ™ ๋“ฑ ๋‹ค์–‘ํ•œ ๋ถˆํ™•์‹ค์„ฑ์„ ๋‚ดํฌํ•œ๋‹ค. ๋ถˆํ™•์‹ค์„ฑ ์ •๋Ÿ‰ํ™”์˜ ์ค‘์š”์„ฑ : ์‹ ๋ขฐํ•  ์ˆ˜ ์žˆ๋Š” ์‹œ๋ฎฌ๋ ˆ์ด์…˜์„ ์œ„ํ•ด์„œ๋Š” ์  ์ถ”์ •๋ฟ ์•„๋‹ˆ๋ผ ์ž˜ ์บ˜๋ฆฌ๋ธŒ๋ ˆ์ด์…˜๋œ ๋ถˆํ™•์‹ค์„ฑ ์ด ํ•„์ˆ˜์ด๋ฉฐ, ์ด๋Š” ํ™œ์„ฑ ํ•™์Šต, ์˜ค๋ฅ˜ ์ „ํŒŒ, ์‹คํ—˜โ€‘์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋น„๊ต ๋“ฑ์— ์ง์ ‘ ํ™œ์šฉ๋œ๋‹ค. ์ „ํ†ต์  ๋ฐฉ๋ฒ•์˜ ํ•œ๊ณ„ : ํ’€ ์•™์ƒ๋ธ”(full ensemble)์€ ๋†’์€ ์ •ํ™•๋„์™€ ๊ฒฌ๊ณ ํ•จ์„ ์ œ๊ณตํ•˜์ง€๋งŒ ํ•™์Šตยท์ถ”๋ก  ๋น„์šฉ์ด n๋ฐฐ ๋กœ ์ฆ๊ฐ€ํ•œ๋‹ค. ๋ผํ”Œ๋ผ์Šค

Physics
No Image

How Well Do Large-Scale Chemical Language Models Transfer to Downstream Tasks?

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ํ™”ํ•™ ์–ธ์–ด ๋ชจ๋ธ(CLM) ์€ SMILESยทSELFIES ๋“ฑ ๋ฌธ์ž์—ด ๊ธฐ๋ฐ˜ ๋ถ„์ž ํ‘œํ˜„์„ NLP ๊ธฐ๋ฒ•์— ์ ์šฉํ•ด ์‚ฌ์ „ํ•™์Šตํ•œ๋‹ค. NLP์—์„œ ์Šค์ผ€์ผ๋ง ๋ฒ•์น™ (๋ชจ๋ธยท๋ฐ์ดํ„ฐยท์—ฐ์‚ฐ๋Ÿ‰ โ†‘ โ†’ ์†์‹ค โ†“, ์„ฑ๋Šฅ โ†‘)์ด ์ž…์ฆ๋œ ๋’ค, ํ™”ํ•™ ๋ถ„์•ผ์—์„œ๋„ ๋™์ผํ•œ ํ˜„์ƒ์ด ๊ธฐ๋Œ€๋˜์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์ตœ๊ทผ LLM ์—ฐ๊ตฌ์—์„œ โ€˜ํ”„๋ฆฌํŠธ๋ ˆ์ธ ์†์‹ค ๊ฐ์†Œ โ‰  ์ „์ด ์„ฑ๋Šฅ ํ–ฅ์ƒโ€™ ์ด๋ผ๋Š” ๋ถ€์ •์  ์ „์ด(negative transfer) ์‚ฌ๋ก€๊ฐ€ ๋ณด๊ณ ๋˜์—ˆ์œผ๋ฉฐ, ์ด๋ฅผ ๊ฒ€์ฆํ•  ํ™”ํ•™ ๋ถ„์•ผ ์—ฐ๊ตฌ๋Š” ๋ถ€์กฑํ–ˆ๋‹ค. 2. ์‹คํ—˜ ์„ค๊ณ„ | ๋ณ€์ˆ˜ | ๋ฒ”์œ„ | ์„ค๋ช… | | | | | | ๋ชจ๋ธ

Computer Science Machine Learning Model
No Image

Hyb-Adam-UM: hybrid ultrametric-aware mtDNA phylogeny reconstruction

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ mtDNA ๋Š” ๋ชจ๊ณ„ ์œ ์ „, ์žฌ์กฐํ•ฉ ์ œํ•œ, ๋†’์€ ๋ณ€์ด์œจ ๋•Œ๋ฌธ์— ์ข… ๊ฐ„ ๊ณ„ํ†ต ๋ถ„์„์— ๋„๋ฆฌ ์“ฐ์ธ๋‹ค. ๊ฑฐ๋ฆฌ ํ–‰๋ ฌ์€ NJ , UPGMA ๋“ฑ ์ „ํ†ต์ ์ธ ๊ณ„ํ†ต์ˆ˜ ์ถ”์ • ์•Œ๊ณ ๋ฆฌ์ฆ˜์˜ ์ง์ ‘ ์ž…๋ ฅ์ด๋ฉฐ, ํ–‰๋ ฌ์˜ ์™„์ „์„ฑยท์ผ๊ด€์„ฑ์ด ๊ฒฐ๊ณผ์— ํฐ ์˜ํ–ฅ์„ ๋ฏธ์นœ๋‹ค. ์ „์ฒด ์Œ์— ๋Œ€ํ•ด NW ์ •๋ ฌ์„ ์ˆ˜ํ–‰ํ•˜๋ฉด O(nยฒยทLยฒ) (L: ์—ผ๊ธฐ ๊ธธ์ด) ์—ฐ์‚ฐ์ด ํ•„์š”ํ•ด, ์ˆ˜์‹ญ~์ˆ˜๋ฐฑ ์ข… ์ˆ˜์ค€์—์„œ๋Š” ์‹ค์šฉ์ ์ด์ง€ ์•Š๋‹ค. 2. ๊ธฐ์กด ๋ฐฉ๋ฒ•์˜ ํ•œ๊ณ„ | ๋ฐฉ๋ฒ• | ํ•ต์‹ฌ ์•„์ด๋””์–ด | ์žฅ์  | ๋‹จ์  | | | | | | | Lowโ€‘rank matrix completion (Sof

Quantitative Biology
No Image

Hybrid Quantum-Classical Optimisation of Traveling Salesperson Problem

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ TSP์˜ ๋‚œ์ด๋„ : NPโ€‘hard ํŠน์„ฑ์œผ๋กœ ์ธํ•ด ๋„์‹œ ์ˆ˜๊ฐ€ ๋Š˜์–ด๋‚ ์ˆ˜๋ก ๊ณ ์ „ ์•Œ๊ณ ๋ฆฌ์ฆ˜๋„ ์‹คํ–‰ ์‹œ๊ฐ„ยท๋ฉ”๋ชจ๋ฆฌ ์š”๊ตฌ๊ฐ€ ๊ธ‰์ฆํ•œ๋‹ค. ์–‘์ž ์ตœ์ ํ™”์˜ ๊ธฐ๋Œ€ : ์–‘์ž ์–ด๋‹๋งยทQAOA ๋“ฑ์€ ํŠน์ • ์กฐํ•ฉ ๋ฌธ์ œ์—์„œ ๊ณ ์ „ ๋Œ€๋น„ ์ด์ ์ด ์žˆ์„ ๊ฐ€๋Šฅ์„ฑ์„ ์ œ์‹œํ–ˆ์ง€๋งŒ, ํ˜„์žฌ NISQ(Nearโ€‘Term Intermediateโ€‘Scale Quantum) ๋””๋ฐ”์ด์Šค๋Š” ๋…ธ์ด์ฆˆ์™€ ํ๋น„ํŠธ ์ˆ˜ ์ œํ•œ์œผ๋กœ ์‹ค์šฉ์ ์ธ ์ˆ˜์ค€์— ๋„๋‹ฌํ•˜์ง€ ๋ชปํ–ˆ๋‹ค. ํ•˜์ด๋ธŒ๋ฆฌ๋“œ ์ ‘๊ทผ : ์–‘์ž์™€ ๊ณ ์ „ ๊ธฐ๋ฒ•์„ ๊ฒฐํ•ฉํ•ด ๊ฐ๊ฐ์˜ ๊ฐ•์ ์„ ์‚ด๋ฆฌ๊ณ  ์•ฝ์ ์„ ๋ณด์™„ํ•˜๋Š” ์ „๋žต์ด ์ตœ๊ทผ ํ™œ๋ฐœํžˆ ์—ฐ๊ตฌ๋˜๊ณ  ์žˆ๋‹ค.

No Image

ICODEN: Ordinary Differential Equation Neural Networks for Interval-Censored Data

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๊ตฌ๊ฐ„ ๊ฒ€์—ด ์€ ์ž„์ƒ ๋ฐฉ๋ฌธ ๊ฐ„๊ฒฉ์ด ๋ถˆ๊ทœ์น™ํ•œ ์žฅ๊ธฐ ์ถ”์  ์—ฐ๊ตฌ(์˜ˆ: ADNI)์—์„œ ํ”ํžˆ ๋ฐœ์ƒํ•œ๋‹ค. ์‚ฌ๊ฑด ๋ฐœ์ƒ ์‹œ์ ์„ ์ •ํ™•ํžˆ ์•Œ ์ˆ˜ ์—†์œผ๋ฏ€๋กœ, ์ „ํ†ต์ ์ธ Kaplanโ€‘MeierยทCoxโ€‘PH์™€ ๊ฐ™์€ ์˜ค๋ฅธ์ชฝ ๊ฒ€์—ด ์ „์šฉ ๋ฐฉ๋ฒ•์„ ๊ทธ๋Œ€๋กœ ์ ์šฉํ•˜๋ฉด ํŽธํ–ฅ์ด ์‹ฌ๊ฐํ•ด์ง„๋‹ค. ๊ณ ์ฐจ์› ๋ฐ”์ด์˜ค๋งˆ์ปค (์ „์žฅ SNP, ๊ณ ํ•ด์ƒ๋„ ์˜์ƒ)์™€ ๊ฐ™์€ ๋ฐ์ดํ„ฐ๊ฐ€ ํญ์ฆํ•˜๋ฉด์„œ, ๊ธฐ์กด ๋ฐ˜์ •๊ทœํ™”(semiparametric)ยท์ •๊ทœํ™”(parametric) ๋ชจ๋ธ์€ ์ฐจ์› ์ €์ฃผ์™€ ๊ณผ์ ํ•ฉ ์œ„ํ—˜์— ์ง๋ฉดํ•œ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด | ์š”์†Œ | ๊ธฐ์กด ๋ฐฉ๋ฒ• | ICODEN | |

Data Computer Science Network Machine Learning
No Image

Impact of Preprocessing on Neural Network-Based RSS/AoA Positioning

| ๊ตฌ๋ถ„ | ๋‚ด์šฉ | ํ‰๊ฐ€ | | | | | | ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ | 5Gยท6G ์‹œ๋Œ€์— ์ •ํ™•ํ•œ ์œ„์น˜์ •๋ณด๊ฐ€ IoT, V2X, UAV ๋“ฑ ๊ณ ์‹ ๋ขฐ ์„œ๋น„์Šค์— ํ•„์ˆ˜์ ์ด๋ฉฐ, RSS๋Š” ์ €๋น„์šฉ, AoA๋Š” ๊ณ ํ•ด์ƒ๋„๋ผ๋Š” ์žฅ์ ์„ ๊ฐ€์ง. ๊ธฐ์กด ์„ ํ˜•ํ™”โ€‘WLS ๋ฐฉ์‹์€ ๋น„์„ ํ˜•์„ฑยท๋‹ค์ค‘๊ฒฝ๋กœยท์žก์Œ์— ์ทจ์•ฝ. | ์ตœ์‹  ํ†ต์‹ ยท์œ„์น˜ํ†ตํ•ฉ ํŠธ๋ Œ๋“œ์™€ ์ž˜ ๋งž์œผ๋ฉฐ, ๋ฌธ์ œ ์ •์˜๊ฐ€ ๋ช…ํ™•ํ•จ. | | ํ•ต์‹ฌ ์•„์ด๋””์–ด | 1) MLP๋ฅผ ์ด์šฉํ•ด RSSโ€‘AoA โ†’ 3D ์ขŒํ‘œ ๋งคํ•‘<br>2) ์ž…๋ ฅ์„ (a) ์›์‹œ ์ธก์ •๊ฐ’ ๊ณผ (b) ์„ ํ˜•ํ™” ๊ธฐ๋ฐ˜ ํŠน์ง• ๋‘ ํ˜•ํƒœ๋กœ ์ œ๊ณตํ•ด ์ „์ฒ˜๋ฆฌ ํšจ๊ณผ๋ฅผ ์ •๋Ÿ‰ํ™” | ๋”ฅ๋Ÿฌ๋‹ ์ ์šฉ

Network Electrical Engineering and Systems Science
No Image

Importance inversion transfer identifies shared principles for cross-domain learning

| ๊ตฌ๋ถ„ | ์ฃผ์š” ๋‚ด์šฉ | ์˜์˜ยท๋น„ํŒ | | | | | | 1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ | ๊ณผํ•™์  ๋ฐœ๊ฒฌ์€ ์„œ๋กœ ๋‹ค๋ฅธ ์Šค์ผ€์ผยท๋ชจ๋‹ฌ๋ฆฌํ‹ฐ๋ฅผ ๊ฐ€์ง„ ์‹œ์Šคํ…œ ๊ฐ„ ๊ณตํ†ต ์›๋ฆฌ ํƒ์ƒ‰์— ์˜์กด.<br> ๊ธฐ์กด ์ „์ด ํ•™์Šต์€ ๋ถ„ํฌ ์ฐจ์ด ์™€ ๋…ธ์ด์ฆˆ ์— ์ทจ์•ฝ, ํŠนํžˆ ์ด์งˆ์ ์ธ ๋„๋ฉ”์ธ์—์„œ๋Š” ์˜๋ฏธ ์žˆ๋Š” ํ‘œํ˜„์„ ์ฐพ๊ธฐ ์–ด๋ ค์›€. | โ€œ๊ณตํ†ต ์›๋ฆฌโ€๋ฅผ ์ฐพ๋Š” ๊ฒƒ์ด ๋‹จ์ˆœํ•œ ๊ธฐ์ˆ ์  ๊ณผ์ œ๊ฐ€ ์•„๋‹ˆ๋ผ ์ด๋ก ์  ์ „์ œ ์ž„์„ ๋ช…ํ™•ํžˆ ํ•จ. ๋‹ค๋งŒ, ์‹ค์ œ ์ ์šฉ ์‚ฌ๋ก€(์˜ˆ: ์—”์ง€๋‹ˆ์–ด๋ง ๊ณ ์žฅ ์˜ˆ์ธก) ์ œ์‹œ๊ฐ€ ๋ถ€์กฑํ•ด ์‹ค์šฉ์„ฑ ํ‰๊ฐ€๊ฐ€ ์ œํ•œ์ . | | 2. Xโ€‘CDTL ํ”„๋ ˆ์ž„์›Œํฌ ์„ค๊ณ„ | ๋„คํŠธ์›Œํฌ๋ฅผ ๋ณดํŽธ์  ๋ฐ์ดํ„ฐ

Computer Science Learning Machine Learning
No Image

In-Situ Analysis of Vibration and Acoustic Data in Additive Manufacturing

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ์˜์˜ FDM ํ”„๋ฆฐํ„ฐ์˜ ํ’ˆ์งˆ ๋ฌธ์ œ ๋Š” ์ง„๋™ยท์Œํ–ฅ ๋“ฑ ๋ฌผ๋ฆฌ์  ๊ต๋ž€์— ํฌ๊ฒŒ ์ขŒ์šฐ๋œ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ๋“ค์€ ์ฃผ๋กœ ์˜คํ”„๋ผ์ธ ๋ถ„์„์— ๋จธ๋ฌผ๋ €์œผ๋‚˜, ๋ณธ ๋…ผ๋ฌธ์€ ์‹ค์‹œ๊ฐ„(inโ€‘situ) ๋ชจ๋‹ˆํ„ฐ๋ง ์„ ๋ชฉํ‘œ๋กœ ํ•˜์—ฌ ์ œ์กฐ ๊ณต์ • ์ค‘ ๊ฒฐํ•จ์„ ์กฐ๊ธฐ์— ํƒ์ง€ํ•œ๋‹ค๋Š” ์ ์—์„œ ์ฐจ๋ณ„์„ฑ์„ ๊ฐ€์ง„๋‹ค. MakerBot Method X ๋Š” ์‚ฐ์—…์šฉ ์ˆ˜์ค€์˜ ๊ณ ์„ฑ๋Šฅ ํ”„๋ฆฐํ„ฐ์ด๋ฏ€๋กœ, ์—ฐ๊ตฌ ๊ฒฐ๊ณผ๊ฐ€ ๋Œ€๊ทœ๋ชจ ์ƒ์‚ฐ ๋ผ์ธ ์— ๋ฐ”๋กœ ์ ์šฉ๋  ๊ฐ€๋Šฅ์„ฑ์ด ํฌ๋‹ค. 2. ์‹คํ—˜ ์„ค๊ณ„ ๋ฐ ๋ฐฉ๋ฒ•๋ก  | ์š”์†Œ | ๋‚ด์šฉ | ํ‰๊ฐ€ | | | | | | ์„ผ์„œ ์„ ํƒ | ADXLโ€‘335 ๊ฐ€์†๋„๊ณ„, SparkFun

Data Electrical Engineering and Systems Science Analysis
No Image

Including Node Textual Metadata in Laplacian-constrained Gaussian Graphical Models

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ ๊ทธ๋ž˜ํ”„ ํ•™์Šต์˜ ํ•„์š”์„ฑ : ๋Œ€๋ถ€๋ถ„์˜ ๊ทธ๋ž˜ํ”„ ์‹ ํ˜ธ ์ฒ˜๋ฆฌยท๊ทธ๋ž˜ํ”„ ์‹ ๊ฒฝ๋ง ๊ธฐ๋ฒ•์€ ์‚ฌ์ „์— ์•Œ๋ ค์ง„ ๊ทธ๋ž˜ํ”„ ๊ตฌ์กฐ๋ฅผ ์ „์ œ๋กœ ํ•˜์ง€๋งŒ, ์‹ค์ œ ๋ฐ์ดํ„ฐ์—์„œ๋Š” ๊ทธ๋ž˜ํ”„๊ฐ€ ๋ฏธ์ง€์ธ ๊ฒฝ์šฐ๊ฐ€ ๋งŽ๋‹ค. ๊ธฐ์กด GGM ํ•œ๊ณ„ : ์ „ํ†ต์ ์ธ ๊ฐ€์šฐ์‹œ์•ˆ ๋งˆ์ฝ”ํ”„ ๋žœ๋ค ํ•„๋“œ(GMRF) ๊ธฐ๋ฐ˜ ๋ฐฉ๋ฒ•์€ ๊ด€์ธก๋œ ์‹ ํ˜ธ๋งŒ์„ ์ด์šฉํ•ด ์ •๋ฐ€ ํ–‰๋ ฌ(precision matrix)์˜ ํฌ์†Œ์„ฑ์„ ํ†ตํ•ด ๊ทธ๋ž˜ํ”„๋ฅผ ์ถ”์ •ํ•œ๋‹ค. ๋ผํ”Œ๋ผ์‹œ์•ˆ ์ œ์•ฝ์„ ์ถ”๊ฐ€ํ•˜๋ฉด ๊ทธ๋ž˜ํ”„๊ฐ€ โ€œ์‹ ํ˜ธ ํ‰ํ™œ(smoothness)โ€์„ ๋งŒ์กฑํ•˜๋„๋ก ๊ฐ•์ œํ•˜์ง€๋งŒ, ๋…ธ๋“œ์— ๋ถ€๊ฐ€์ ์ธ ํ…์ŠคํŠธ ์„ค๋ช… ๋“ฑ ๋ฉ”ํƒ€๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•˜์ง€ ๋ชปํ•œ๋‹ค.

Data Statistics Machine Learning Model
No Image

Inequalities For The Growth Of Rational Functions With Prescribed Poles

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ๋‹คํ•ญ์‹ ์„ฑ์žฅ ๋ถˆํ‰๋“ฑ : ์ตœ๋Œ€ ๋ชจ๋“ˆ๋Ÿฌ์Šค ์›๋ฆฌ์™€ Varga, Rivlin, Aziz, Kumarโ€‘Milovanoviฤ‡, Dhankharโ€‘Kumar ๋“ฑ์˜ ์ž‘์—…์„ ํ†ตํ•ด (|z| 1) ์œ„์—์„œ (|f(z)|) ์™€ (|f'(z)|) ์‚ฌ์ด์˜ ๊ด€๊ณ„๊ฐ€ ๊ณ„์ˆ˜์— ๋”ฐ๋ผ ์–ด๋–ป๊ฒŒ ๊ฐ•ํ™”๋  ์ˆ˜ ์žˆ๋Š”์ง€๊ฐ€ ์ฒด๊ณ„ํ™”๋˜์—ˆ๋‹ค. ์œ ๋ฆฌํ•จ์ˆ˜๋กœ์˜ ํ™•์žฅ : LiยทMohapatraยทRodriguez, ๊ทธ๋ฆฌ๊ณ  ์ตœ๊ทผ Rather et al.์ด ๋‹คํ•ญ์‹ ๊ฒฐ๊ณผ๋ฅผ ๊ทน์ ์ด (|beta j|>1) ์ธ ์œ ๋ฆฌํ•จ์ˆ˜๋กœ ์ผ๋ฐ˜ํ™”ํ–ˆ์ง€๋งŒ, ๊ณ„์ˆ˜ ์ •๋ณด๋ฅผ ์ถฉ๋ถ„ํžˆ ํ™œ์šฉํ•˜์ง€ ๋ชป

Mathematics
No Image

Intellicise Wireless Networks Meet Agentic AI: A Security and Privacy Perspective

1. ๋…ผ๋ฌธ์˜ ๊ตฌ์กฐ์™€ ํ•ต์‹ฌ ๊ธฐ์—ฌ | ๋ฒˆํ˜ธ | ํ•ต์‹ฌ ๋‚ด์šฉ | ์˜์˜ | | | | | | โ‘  | ์ธํ…”๋ฆฌ์‹œ์ฆˆ ๋ฌด์„  ๋„คํŠธ์›Œํฌ์™€ ์—์ด์ „ํŠธ AI ๊ฐœ๋… ์ •์˜ | ๋‘ ๊ธฐ์ˆ ์˜ ๊ธฐ๋ณธ ์›๋ฆฌ๋ฅผ ๋ช…ํ™•ํžˆ ๊ตฌ๋ถ„ํ•˜๊ณ , ์œตํ•ฉ ๊ฐ€๋Šฅ์„ฑ์„ ์ œ์‹œ | | โ‘ก | ์—์ด์ „ํŠธ AI๊ฐ€ ์ œ๊ณตํ•˜๋Š” ์ฃผ์š” ์žฅ์  (ํ”„๋ฆฌ์•กํ‹ฐ๋ธŒ ๋ฐฉ์–ด, ๋™์  ํ˜‘์—…, ์ง€์† ํ•™์Šต) | ๊ธฐ์กด ์ •์  ๋ฐฉ์–ด ์ฒด๊ณ„์™€ ์ฐจ๋ณ„ํ™”๋˜๋Š” ์ž์œจยท์ง€์† ๊ฐ€๋Šฅํ•œ ๋ฐฉ์–ด ๋ฉ”์ปค๋‹ˆ์ฆ˜์„ ๊ฐ•์กฐ | | โ‘ข | ๋ณด์•ˆยทํ”„๋ผ์ด๋ฒ„์‹œ ์œ„ํ—˜ ๋ถ„๋ฅ˜ (์„ผ์‹ฑยท๋น”ํฌ๋ฐยท๋ผ๋””์˜ค๋งตยท์Šคํ…Œ๊ฐ€๋…ธ๊ทธ๋ž˜ํ”ผ ๋“ฑ) | ๊ฐ ๊ณ„์ธตยท๋ชจ๋“ˆ๋ณ„ ์œ„ํ˜‘์„ ์ฒด๊ณ„ํ™”ํ•˜์—ฌ ์—ฐ๊ตฌยท์‹ค๋ฌด ์ ์šฉ์„ฑ์„ ๋†’์ž„ | | โ‘ฃ

Computer Science Network Cryptography and Security
No Image

Joint analysis for multivariate longitudinal and event time data with a change point anchored at interval-censored event time

| ๊ตฌ๋ถ„ | ๋‚ด์šฉ | ํ‰๊ฐ€ยท๋น„ํŒ | | | | | | ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ | ๊ธฐ์กด ๊ณต๋™ ๋ชจ๋ธ์€ ๋Œ€๋ถ€๋ถ„ ์šฐ์ธก ๊ฒ€์—ด(rightโ€‘censored) ์‚ฌ๊ฑด์‹œ๊ฐ„์— ์ดˆ์ ์„ ๋งž์ถ”์—ˆ์œผ๋ฉฐ, ๊ตฌ๊ฐ„ ๊ฒ€์—ด(event interval)๊ณผ ๋ณ€๊ณก์ (changeโ€‘point) ์„ ๋™์‹œ์— ๋‹ค๋ฃจ์ง€๋Š” ๋ชปํ–ˆ๋‹ค. HD์™€ ๊ฐ™์ด ๋ฐœ๋ณ‘ ์‹œ์ ์ด ์ •ํ™•ํžˆ ๊ด€์ธก๋˜์ง€ ์•Š๋Š” ์ƒํ™ฉ์—์„œ, ์ธ์ง€ยท์šด๋™ ๋ฐ”์ด์˜ค๋งˆ์ปค์˜ ์ „ํ›„ ๋ณ€ํ™”๋ฅผ ๋™์‹œ์— ๋ถ„์„ํ•  ํ•„์š”๊ฐ€ ์žˆ๋‹ค. | ์—ฐ๊ตฌ ๋™๊ธฐ๊ฐ€ ๋ช…ํ™•ํ•˜๊ณ , ํ˜„์žฌ ๋ฐฉ๋ฒ•๋ก ์˜ ํ•œ๊ณ„๋ฅผ ์ •ํ™•ํžˆ ์งš๊ณ  ์žˆ๋‹ค. ํŠนํžˆ โ€œ์ธ์ง€ ์ €ํ•˜ โ†’ ๋ฐœ๋ณ‘ โ†’ ์ธ์ง€ ์•…ํ™”โ€๋ผ๋Š” ์–‘๋ฐฉํ–ฅ ์ธ๊ณผ ๊ตฌ์กฐ๋ฅผ ๋ชจ๋ธ๋ง

Data Statistics Analysis
No Image

Joint beamforming and mode optimization for multi-functional STAR-RIS-aided integrated sensing and communication networks

| ๊ตฌ๋ถ„ | ํ•ต์‹ฌ ๋‚ด์šฉ | ๊ธฐ์ˆ ์  ์˜์˜ | ๊ธฐ๋Œ€ ํšจ๊ณผ | | | | | | | ์‹œ์Šคํ…œ ๋ชจ๋ธ | BS: Mโ€‘์•ˆํ…Œ๋‚˜ ULA <br> STARโ€‘RIS: Nร—N UPA <br> ์‹ค๋‚ด(K T)ยท์‹ค์™ธ(K R) ์‚ฌ์šฉ์ž ๊ฐ๊ฐ ๋‹จ์ผ ์•ˆํ…Œ๋‚˜ <br> ์ง์ ‘ BSโ€‘์‚ฌ์šฉ์ž ๋งํฌ ๋ฌด์‹œ (์ฐจํยท๊ณ ์ฃผํŒŒ ์†์‹ค) | ์ „ํŒŒ ์†์‹ค์ด ํฐ mmWave/THz ํ™˜๊ฒฝ์„ ํ˜„์‹ค์ ์œผ๋กœ ๋ฐ˜์˜, RISโ€‘๊ฒฝ์œ  ์ „ํŒŒ๋งŒ ๊ณ ๋ คํ•จ์œผ๋กœ ์„ค๊ณ„ ๋ณต์žก๋„ ๊ฐ์†Œ | ์‹ค์ œ 6Gยท์Šค๋งˆํŠธ ์‹œํ‹ฐ ์‹œ๋‚˜๋ฆฌ์˜ค์— ์ ํ•ฉ | | ๋‘ ๋‹จ๊ณ„ ํ”„๋กœํ† ์ฝœ | โ‘  ์ค€๋น„ ๋‹จ๊ณ„(ฮทยทT) : ์ด์ „ ์Šฌ๋กฏ์—์„œ ์–ป์€ DOA ์ถ”์ •๊ฐ’ ์‚ฌ์šฉ, ๋™

Network Electrical Engineering and Systems Science
No Image

Joint Modeling of Longitudinal EHR Data with Shared Random Effects for Informative Visiting and Observation Processes

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ ๋‘ ๋‹จ๊ณ„์˜ ์ •๋ณด ์†์‹ค : 1) Informative Presence (IP) โ€“ ํ™˜์ž์˜ ๊ฑด๊ฐ• ์ƒํƒœ๊ฐ€ ๋ฐฉ๋ฌธ ๋นˆ๋„์— ์˜ํ–ฅ์„ ๋ฏธ์ณ ๋ฐฉ๋ฌธ ์ž์ฒด๊ฐ€ ์„ ํƒ์ ์ž„. 2) Informative Observation (IO) โ€“ ๋ฐฉ๋ฌธ์ด ์ด๋ฃจ์–ด์ ธ๋„ ํŠน์ • ๋ฐ”์ด์˜ค๋งˆ์ปค๊ฐ€ ์ธก์ •๋ ์ง€๋Š” ๋˜ ๋‹ค๋ฅธ ์ž„์ƒ ํŒ๋‹จ์— ๋”ฐ๋ผ ๋‹ฌ๋ผ์ง. ๊ธฐ์กด ํ†ต๊ณ„ ๋ฐฉ๋ฒ•์€ ์ฃผ๋กœ IP๋งŒ์„ ๊ณ ๋ คํ•˜๊ฑฐ๋‚˜, ๊ด€์ธก์ด ํ•ญ์ƒ ์ด๋ฃจ์–ด์ง„๋‹ค๊ณ  ๊ฐ€์ •ํ•ด IO๋ฅผ ๋ฌด์‹œํ•œ๋‹ค. ์ด๋Š” MNAR (Missing Not At Random) ์ƒํ™ฉ์„ ์ œ๋Œ€๋กœ ๋ฐ˜์˜ํ•˜์ง€ ๋ชปํ•œ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด โ€“ ๊ณต์œ  ๊ฐ€

Data Statistics Model
No Image

Kalman Filtering Based Flight Management System Modeling for AAM Aircraft

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ AAM ํŠน์„ฑ : ์งง์€ ๊ฑฐ๋ฆฌยท๋„์‹ฌ ์ €๊ณ ๋„ ๋น„ํ–‰, ๊ณ ์ž์œจ์„ฑ, ATC ๋ถ€ํ•˜ ์ตœ์†Œํ™” ๋ชฉํ‘œ. ๊ธฐ์กด ๋ฌธ์ œ : NAS๊ฐ€ AAM ์ฆ๊ฐ€์— ๋Œ€์‘ํ•˜๊ธฐ ์–ด๋ ค์šฐ๋ฉฐ, ์ „๋žต์  ๋น„ํ–‰๊ณ„ํš ์„œ๋น„์Šค๊ฐ€ ์—†์œผ๋ฉด ATC์— ๊ณผ๋„ํ•œ ์ž‘์—… ๋ถ€ํ•˜๊ฐ€ ๋ฐœ์ƒ. ๋ถˆํ™•์‹ค์„ฑ ๋ชจ๋ธ๋ง์˜ ์ค‘์š”์„ฑ : RTA(Required Time of Arrival) ๋ณ€๋™์„ฑ์„ ์ •๋Ÿ‰ํ™”ํ•ด์•ผ ํ•ญ๊ณต๊ธฐ ๊ฐ„ ์ถฉ๋Œ ์œ„ํ—˜, ์šฉ๋Ÿ‰ ์ดˆ๊ณผ ๋“ฑ์„ ์‚ฌ์ „ ํŒ๋‹จ ๊ฐ€๋Šฅ. 2. ๊ธฐ์กด ์—ฐ๊ตฌ์™€์˜ ์ฐจ๋ณ„์  | ์—ฐ๊ตฌ | ๋ถˆํ™•์‹ค์„ฑ ๋ชจ๋ธ | ์ฃผ์š” ํ•œ๊ณ„ | | | | | |

Robotics Computer Science System Model
No Image

Large elements and advanced beamformers for increased field of view in 2-D ultrasound matrix arrays

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์š”์†Œ ์ˆ˜์™€ ์ฑ„๋„ ์ œํ•œ : 32ร—32(1024) ์š”์†Œ ์–ด๋ ˆ์ด๋Š” ํ˜„์žฌ ๋Œ€๋ถ€๋ถ„์˜ ์ดˆ์ŒํŒŒ ์‹œ์Šคํ…œ์ด ์ง€์›ํ•˜์ง€ ๋ชปํ•˜๋Š” ์ฑ„๋„ ์ˆ˜๋ฅผ ์š”๊ตฌํ•œ๋‹ค. ๋ฉ€ํ‹ฐํ”Œ๋ ‰์‹ฑ์€ ํ”„๋ ˆ์ž„ ๋ ˆ์ดํŠธ๋ฅผ ํฌ๊ฒŒ ์ €ํ•˜์‹œํ‚จ๋‹ค. ์‹œ์•ผ ์ œํ•œ : ์š”์†Œ ํ”ผ์น˜๊ฐ€ ฮป ์ดํ•˜๋กœ ์„ค๊ณ„๋ผ์•ผ ๊ทธ๋ ˆ์ดํŒ… ๋กœ๋ธŒ๋ฅผ ์–ต์ œํ•˜์ง€๋งŒ, ์ด๋Š” ์–ดํผ์ฒ˜ ๋ฉด์ ์„ ์ œํ•œํ•ด FOV๊ฐ€ ์ž‘์•„์ง„๋‹ค. ๋Œ€์•ˆ ์ ‘๊ทผ๋ฒ• : ๋งˆ์ดํฌ๋กœโ€‘๋น”ํฌ๋จธ, ์ŠคํŒŒ์Šค ์–ด๋ ˆ์ด, RCA ์–ด๋ ˆ์ด ๋“ฑ์€ ์š”์†Œ ์ˆ˜๋ฅผ ์ค„์ด์ง€๋งŒ ๊ฐ๊ฐ ํ•ด์ƒ๋„ยทSNR ์ €ํ•˜ ํ˜น์€ ์ œํ•œ๋œ ์Šคํ‹ฐ์–ด๋ง ์ด๋ผ๋Š” ๋‹จ์ ์„ ๊ฐ€์ง„๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด ์š”์†Œ ํฌ๊ธฐ ํ™•๋Œ€ โ†’ ํฐ ์š”์†Œ๋Š” ์ง์ ‘

Electrical Engineering and Systems Science

< Category Statistics (Total: 2829) >

Electrical Engineering and Systems Science
100
General
731
General Relativity
22
HEP-EX
17
HEP-LAT
3
HEP-PH
39
HEP-TH
19
MATH-PH
36
NUCL-EX
2
NUCL-TH
5
Quantum Physics
41

Start searching

Enter keywords to search articles

โ†‘โ†“
โ†ต
ESC
โŒ˜K Shortcut