DeveLife

Written by@[Anthony min]
Develop product, Develop life, Develop myself.

GitHubFacebookInstagram

์ข‹์€ PR์— ๋Œ€ํ•œ ๋‹จ์ƒ ๐Ÿค”

๋ณธ ๊ฒŒ์‹œ๋ฌผ์€ ํ•˜์–€๋งˆ์ธ๋“œ ๊ธฐ์ˆ ๋ธ”๋กœ๊ทธ์— ๋ณธ์ธ์ด ๊ธฐ๊ณ ํ•œ ๊ธ€์„ ์˜ฎ๊ฒจ ์ž‘์„ฑํ–ˆ์Šต๋‹ˆ๋‹ค.

Main-image

์ด ๊ธ€์„ ์“ฐ๊ฒŒ ๋œ ์ด์œ 

์ด ๋‚ด์šฉ์€ ์˜์–ด๋กœ ์ž‘์„ฑํ•˜๊ณ  ์‹ถ์—ˆ์œผ๋‚˜, ์ข‹์ง€ ๋ชปํ•œ ์˜์–ด ์‹ค๋ ฅ๋•Œ๋ฌธ์— ๋‚จ๊ธฐ๊ณ ์ž ํ•˜๋Š” ์˜๋ฏธ๊ฐ€ ์™œ๊ณก๋  ์ˆ˜๋„ ์žˆ๊ณ , ์˜ฌ๋ฐ”๋ฅธ ํ‘œํ˜„๊ณผ ์–ดํœ˜๋ฅผ ์ฐพ๊ธฐ ์œ„ํ•œ ์‹œ๊ฐ„์„ ๋“ค์ด๊ธฐ ๋ณด๋‹ค๋Š” ๊ธ€์„ ๋” ์ž˜ ์ž‘์„ฑํ•˜๋Š” ๋ฐ ์ง‘์ค‘ํ•˜๊ณ ์ž ํ•œ๊ตญ์–ด๋กœ ์ž‘์„ฑํ•˜๋ ค๊ณ  ํ•œ๋‹ค.

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 08

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
์ฝ”๋“œ ์ถœ์ฒ˜

Lecture 8. Deep learning Basic : History

์ด๋ฒˆ ๊ฐ•์˜๋ถ€ํ„ฐ ๋ณธ๊ฒฉ์ ์œผ๋กœ Deep neural network์— ๊ด€ํ•œ ์ด์•ผ๊ธฐ๋ฅผ ๋‹ค๋ฃฐ ๊ฒƒ์ด๋‹ค. ์ด๋ฒˆ ์žฅ์€ ์ด ๋”ฅ๋Ÿฌ๋‹์ด๋ผ๋Š” ์•„์ด๋””์–ด๊ฐ€ ์–ด๋–ป๊ฒŒ ์‹œ์ž‘๋˜์—ˆ๋Š”์ง€, ๊ทธ๋ฆฌ๊ณ  ์–ด๋– ํ•œ ๋ฌธ์ œ๊ฐ€ ์žˆ์—ˆ๊ณ  ๊ทธ ๋ฌธ์ œ๋“ค์„ ์ธ๋ฅ˜๊ฐ€ ์–ด๋–ป๊ฒŒ ํ•ด๊ฒฐํ•ด์™”๋Š”์ง€์— ๋Œ€ํ•ด์„œ ์ˆ˜ํ•™์ ,์ปดํ“จํ„ฐ์ ์ธ ์ž์„ธํ•œ ๋‚ด์šฉ์„ ๋ฐฐ์ œํ•˜๊ณ โ€ฆ

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 07

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
์ฝ”๋“œ ์ถœ์ฒ˜

Lecture 7. Application & Tips

์ด๋ฒˆ ๊ฐ•์˜๋Š” Machine Learning Algorithm์„ ์‹ค์ œ๋กœ ์ ์šฉํ•จ์— ์žˆ์–ด์„œ ์ค‘์š”ํ•œ ๋ช‡ ๊ฐ€์ง€ ํŒ๋“ค์— ๋Œ€ํ•ด์„œ ์•Œ์•„ ๋ณผ ๊ฒƒ์ด๋‹ค.

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 06

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
์ฝ”๋“œ ์ถœ์ฒ˜

Lecture 6. Softmax Classification

์—ฌ๋Ÿฌ ๊ฐœ์˜ ํด๋ž˜์Šค๊ฐ€ ์žˆ์„ ๋•Œ, ๊ทธ ๊ฒƒ์„ ์˜ˆ์ธกํ•˜๋Š” ๋ฐฉ๋ฒ•์„ Multinomial Classification
์ด๋ผ๊ณ  ํ•˜๋ฉฐ, ๊ทธ ์ค‘์— ๊ฐ€์žฅ ๋งŽ์ด ์‚ฌ์šฉ๋˜๋Š” Softmax Classification์— ๋Œ€ํ•˜์—ฌ ๋ฐฐ์›Œ๋ณด๋„๋ก ํ•œ๋‹ค.

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 05

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
์ฝ”๋“œ ์ถœ์ฒ˜

Lecture 5. Logistic Classification

Logistic classification์€ classification algorithm๋“ค ์ค‘์—์„œ ์ •ํ™•๋„๊ฐ€ ๋†’์€ ๊ฒƒ์œผ๋กœ ์•Œ๋ ค์ ธ ์žˆ๋‹ค.
๋•Œ๋ฌธ์— ์‹ค์ œ ๋ฌธ์ œ์—๋„ ๋ฐ”๋กœ ์ ์šฉํ•ด๋ณผ ์ˆ˜ ์žˆ๊ณ , Neural Network์™€ Deep learning์„ ์ดํ•ดํ•˜๋Š” ๋ฐ ์ค‘์š”ํ•œ ์ปดํฌ๋„ŒํŠธ์ด๋‹ค.

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 04

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 03

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
์ฝ”๋“œ ์ถœ์ฒ˜

์ด๋ก  ๋ถ€๋ถ„์— ํ•ด๋‹นํ•˜๋Š” ๋‚ด์šฉ์€ ์ˆ˜์‹์„ ํ‘œํ˜„ํ•˜๋Š” ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ์ž˜ ๋‹ค๋ฃฐ์ค„ ๋ชฐ๋ผ์„œ ์‚ฌ์ง„์œผ๋กœ ์ฐ์„์ˆ˜๋ฐ–์— ์—†์–ด ๊ธธ์–ด์ง„๋‹คโ€ฆ

Lecture 3. Linear Regression cost ํ•จ์ˆ˜ ์ตœ์†Œํ™”

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 02-2

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
์ฝ”๋“œ ์ถœ์ฒ˜

์„น์…˜ 2 ์‹ค์Šต (Linear regression ๊ตฌํ˜„)

์‹ค์Šตํ•˜๊ธฐ ์ „์—, Hypothesis์™€ Cost function ๋ณต์Šต!

hypo

Hypothesis๋ž€, ์ฃผ์–ด์ง„ x์— ๋Œ€ํ•˜์—ฌ ์šฐ๋ฆฌ๊ฐ€ ์˜ˆ์ธก์„ ์–ด๋–ป๊ฒŒ ํ•  ๊ฒƒ์ธ๊ฐ€ ๋ผ๋Š” ๊ฒƒ์„ ๋งํ•œ๋‹ค. ์ด๋Š” W์™€ x ์˜ ๊ณฑ, ๊ทธ๋ฆฌ๊ณ  bias์™€์˜ ํ•ฉ์œผ๋กœ ๊ฒฐ์ •๋œ๋‹ค. ๊ทธ๋ฆฌ๊ณ  ์ด ๊ฒƒ์„ ์–ผ๋งˆ๋‚˜ ์ž˜ ์˜ˆ์ธกํ–ˆ๋Š”๊ฐ€๋ฅผ ์ธก์ •ํ•˜๊ธฐ ์œ„ํ•ด ์˜ˆ์ธก ๊ฐ’๊ณผ ์ฐธ ๊ฐ’์˜ ์ฐจ์ด์˜ ์ œ๊ณฑ์„ ์ „์ฒด ๋ฐ์ดํ„ฐ์˜ ๊ฐœ์ˆ˜๋กœ ๋‚˜๋ˆˆ ํ‰๊ท ์ด ๋ฐ”๋กœ

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 02-1

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
์ฝ”๋“œ ์ถœ์ฒ˜

์ง€๋‚œ ์‹œ๊ฐ„์— ๋ชปํ•œ TensorFlow ์‹ค์Šต

์„น์…˜ 1 ์‹ค์Šต (๊ธฐ๋ณธ์ ์ธ operations)

import tensorflow as tf

hello = tf.constant("Hello, TensorFlow!")

sess = tf.Session()

print(sess.run(hello))

์œ„์˜ ์ฝ”๋“œ๋Š” ์šฐ๋ฆฌ๊ฐ€ ํ”„๋กœ๊ทธ๋ž˜๋ฐ์„ ๋ฐฐ์šฐ๋ฉด์„œ ๊ฐ€์žฅ ํ”ํžˆ ์•Œ๊ณ , ๊ฐ€์žฅ ๊ธฐ๋ณธ์ ์ธ hello world๋ฅผ ํ…์„œํ”Œ๋กœ์—์„œ ์‹คํ–‰ํ•˜๋Š” ์ฝ”๋“œ์ด๋‹ค. ์ •๋ง ๊ฐ„๋‹จํ•˜์ง€๋งŒ, ๋‚˜ ์Šค์Šค๋กœ๋„ ํ…์„œํ”Œ๋กœ์™€ ๋จธ์‹ ๋Ÿฌ๋‹์„ ์ฒ˜์Œ ์ ‘ํ•˜๊ธฐ ๋•Œ๋ฌธ์—, ํ•˜๋‚˜ํ•˜๋‚˜ ์‚ดํŽด๋ณด์ž๋ฉด tensorflow ๋ฅผ import ํ•˜์—ฌ tf๋ผ๋Š” ์ด๋ฆ„์œผ๋กœ ์‚ฌ์šฉํ•˜๊ธฐ๋กœ ํ–ˆ์—ˆ๋‹ค.

[course] ๋ชจ๋‘๋ฅผ ์œ„ํ•œ ๋”ฅ๋Ÿฌ๋‹ ๊ฐ•์ขŒ 01

์ด ํฌ์ŠคํŒ…์€ ์ธํ”„๋Ÿฐ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ฐ•์ขŒ๋ฅผ ์ˆ˜๊ฐ•ํ•˜๋ฉฐ ๊ณต๋ถ€ํ•œ ๋‚ด์šฉ์„ ์ •๋ฆฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

์‚ฌ์‹ค ์ด ๊ฐ•์ขŒ์— ๋Œ€ํ•ด์„œ ์ ‘ํ•œ ์ง€๋Š” ๊ฝค ์˜ค๋žœ ์‹œ๊ฐ„์ด ์ง€๋‚ฌ๋Š”๋ฐ, ์ˆ˜๊ฐ•ํ•˜๊ฒŒ ๋œ ๊ฒƒ๋„, ์ด ๋ถ„์•ผ๋ฅผ ์ ‘ํ•˜๊ฒŒ ๋œ ๊ฒƒ๋„ ์ง€๊ธˆ์— ์™€์„œ์•ผ ๋ผ์„œ ์•„์‰ฌ์šด ๊ฐ์ด ์žˆ๋‹ค.

์„น์…˜1. ๋จธ์‹ ๋Ÿฌ๋‹์˜ ๊ฐœ๋…๊ณผ ์šฉ์–ด

๋จธ์‹ ๋Ÿฌ๋‹์ด๋ž€ ๋ฌด์—‡์ธ๊ฐ€?