728x90
๋ฐ˜์‘ํ˜•

GRU 2

[๋ถ€์ŠคํŠธ์บ ํ”„ AI Tech 3๊ธฐ Pre-Course ์ˆ˜๋ฃŒ] ์ธ๊ณต์ง€๋Šฅ ๋ณธ๊ฒฉ ํƒ๊ตฌ : RNN - Sequential Models

๋ชจ๋“  ๊ฒŒ์‹œ๋ฌผ์€ macOS Monterey 12.0.1 ๋ฒ„์ „ ๊ธฐ์ค€์œผ๋กœ ์ž‘์„ฑํ•˜์˜€์Šต๋‹ˆ๋‹ค. ๋ถ€์ŠคํŠธ์บ ํ”„ AI Tech 3๊ธฐ๋ฅผ ์œ„ํ•œ Pre-Course ๋ฅผ ํ† ๋Œ€๋กœ ์ž‘์„ฑํ•˜์˜€์Šต๋‹ˆ๋‹ค. Sequential Model Sequential Data๋ฅผ ์ฒ˜๋ฆฌํ•˜๋Š” ๊ฐ€์žฅ ํฐ ์–ด๋ ค์›€์€ ์ง€๊ธˆ๊นŒ์ง€๋Š” ์šฐ๋ฆฌ๊ฐ€ ์–ป๊ณ  ์‹ถ์€ ๊ฒƒ์ด ํ•˜๋‚˜์˜ Label์ด์—ˆ๋Š”๋ฐ Sequential Data๋Š” ๊ธธ์ด๊ฐ€ ์ œ๊ฐ๊ฐ์ด๋‹ค. ๋ฐ›์•„๋“ค์ผ ์ž…๋ ฅ์˜ ์ฐจ์›์ด ๊ณ ์ •๋˜์–ด ์žˆ์ง€ ์•Š๋‹ค๋Š” ๊ฒƒ์ด๋‹ค. ๋ช‡๊ฐœ์˜ ๋‹จ์–ด, ์Œ์ ˆ, ์ด๋ฏธ์ง€๊ฐ€ ๋“ค์–ด์˜ฌ ์ง€ ๋ชจ๋ฅด๋ฏ€๋กœ ๊ทธ ํฌ๊ธฐ์— ์ƒ๊ด€์—†์ด ๋™์ž‘ํ•  ๋ชจ๋ธ์ด ํ•„์š”ํ•˜๋‹ค. ์ด์ „์— ์–ด๋–ค ๋ง์ด ๋“ค์–ด์™”์„ ๋•Œ ๋‹ค์Œ ๋ง์ด ๋ฌด์—‡์ผ์ง€ ์˜ˆ์ธกํ•˜๋Š” ๋ฌธ์ œ. ์‹œ๊ฐ„์ด ์ง€๋‚ ์ˆ˜๋ก ๊ณ ๋ คํ•ด์•ผ ํ•˜๋Š” ์ง€๋‚œ ์‹œ๊ฐ„์˜ ๋ฐ์ดํ„ฐ๋“ค์ด ๋Š˜์–ด๋‚œ๋‹ค. ์ด๋ฅผ ๊ฐ€์žฅ ๊ฐ„๋‹จํžˆ ๋งŒ๋“œ๋Š” ๊ฒƒ์€ Fixed Time Span. ๊ณผ๊ฑฐ์˜ ..

[๋ถ€์ŠคํŠธ์บ ํ”„ AI Tech 3๊ธฐ Pre-Course] ํ•ต์‹ฌ๋งŒ ๋‹ด์€ AI๋ฅผ ์œ„ํ•œ ์ˆ˜ํ•™ ์ง€์‹ CNN, RNN ์ฒซ๊ฑธ์Œ

๋ชจ๋“  ๊ฒŒ์‹œ๋ฌผ์€ macOS Monterey 12.0.1 ๋ฒ„์ „ ๊ธฐ์ค€์œผ๋กœ ์ž‘์„ฑํ•˜์˜€์Šต๋‹ˆ๋‹ค. ๋ถ€์ŠคํŠธ์บ ํ”„ AI Tech 3๊ธฐ๋ฅผ ์œ„ํ•œ Pre-Course ๋ฅผ ํ† ๋Œ€๋กœ ์ž‘์„ฑํ•˜์˜€์Šต๋‹ˆ๋‹ค. Convolution & CNN [ ๊ฐ•์˜ ์†Œ๊ฐœ ] Convolution ์—ฐ์‚ฐ๊ณผ ๋‹ค์–‘ํ•œ ์ฐจ์›์—์„œ์˜ ์—ฐ์‚ฐ๋ฐฉ๋ฒ•์„ ๊ณต๋ถ€ํ•ฉ๋‹ˆ๋‹ค. Convolution ์—ฐ์‚ฐ์˜ ์—ญ์ „ํŒŒ์— ๋Œ€ํ•ด ๊ณต๋ถ€ํ•ฉ๋‹ˆ๋‹ค. Convolution์€ ์ด๋ฏธ์ง€๋‚˜ ์˜์ƒ ์ฒ˜๋ฆฌ์— ์žˆ์–ด์„œ ์ž์ฃผ ํ™œ์šฉ๋˜๋Š” ์—ฐ์‚ฐ์ด๋‹ค. Fully Connected Network์— ๋น„ํ•ด Convolutional Neural Network๊ฐ€ ๊ฐ€์ง€๋Š” ์žฅ์ ์„ ์ดํ•ดํ•˜๊ณ  ๊ทธ ์—ญ์ „ํŒŒ ์—ฐ์‚ฐ์— ๋Œ€ํ•ด ๊ณต๋ถ€ํ•œ๋‹ค. https://setosa.io/ev/image-kernels/ Image Kernels explained visually An i..

728x90
๋ฐ˜์‘ํ˜•