Potato
์•ˆ๋…•ํ•˜์„ธ์š”, ๊ฐ์žก๋‹ˆ๋‹ค?๐Ÿฅ” ^___^ ๐Ÿ˜บ github ๋ฐ”๋กœ๊ฐ€๊ธฐ ๐Ÿ‘‰๐Ÿป
๋ฐ˜์‘ํ˜•

AI study 35

[๋”ฅ๋Ÿฌ๋‹] ์†์‹คํ•จ์ˆ˜ (loss function) ์ข…๋ฅ˜ ๋ฐ ๊ฐ„๋‹จ ์ •๋ฆฌ (feat. keras & pytorch)

์‹œ์ž‘ํ•˜๋ฉฐ ๋”ฅ๋Ÿฌ๋‹๋ชจ๋ธ ๊ตฌ์ถ• ๋„์ค‘ ์†์‹คํ•จ์ˆ˜ ๋•Œ๋ฌธ์— ์˜ค๋ฅ˜๊ฐ€ ๋‚ฌ๋‹ค. ์•„๋งˆ ์†์‹คํ•จ์ˆ˜์™€ ํ™œ์„ฑํ™” ํ•จ์ˆ˜์˜ ์กฐํ•ฉ์ด ๋งž์ง€ ์•Š์•˜๋˜ ๊ฒƒ ๊ฐ™๋‹ค. ์ผ๋‹จ ๊ทธ๋ž˜์„œ ์ด๋Œ€๋กœ๋Š” ์•ˆ๋˜๊ฒ ๋‹ค ์‹ถ์–ด์„œ ์ž์„ธํ•œ ์ˆ˜์‹๊นŒ์ง€๋Š” ์•„๋‹ˆ๋”๋ผ๋„ ์–ธ์ œ, ์–ด๋–ป๊ฒŒ, ๋ฌด์Šจ ์ข…๋ฅ˜์˜ ์†์‹คํ•จ์ˆ˜๊ฐ€ ์žˆ๋Š”์ง€ ์‚ดํŽด๋ณด๊ธฐ๋กœ ํ•œ๋‹ค!!! ์•„์ž์•„์ž ํ™”์ดํŒ… โ‹Œเผผ •ฬ€ โŒ‚ •ฬ เผฝโ‹‹ ์†์‹ค ํ•จ์ˆ˜(loss function) ๋ž€? ๋จธ์‹ ๋Ÿฌ๋‹ ํ˜น์€ ๋”ฅ๋Ÿฌ๋‹ ๋ชจ๋ธ์˜ ์ถœ๋ ฅ๊ฐ’๊ณผ ์‚ฌ์šฉ์ž๊ฐ€ ์›ํ•˜๋Š” ์ถœ๋ ฅ๊ฐ’์˜ ์˜ค์ฐจ๋ฅผ ์˜๋ฏธ ์†์‹คํ•จ์ˆ˜๋Š” ์ •๋‹ต(y)์™€ ์˜ˆ์ธก(^y)๋ฅผ ์ž…๋ ฅ์œผ๋กœ ๋ฐ›์•„ ์‹ค์ˆซ๊ฐ’ ์ ์ˆ˜๋ฅผ ๋งŒ๋“œ๋Š”๋ฐ, ์ด ์ ์ˆ˜๊ฐ€ ๋†’์„์ˆ˜๋ก ๋ชจ๋ธ์ด ์•ˆ์ข‹์€ ๊ฒƒ ์†์‹คํ•จ์ˆ˜์˜ ํ•จ์ˆ˜๊ฐ’์ด ์ตœ์†Œํ™” ๋˜๋„๋ก ํ•˜๋Š” ๊ฐ€์ค‘์น˜(weight)์™€ ํŽธํ–ฅ(bias)๋ฅผ ์ฐพ๋Š” ๊ฒƒ์ด ๋ชฉํ‘œ ํŒŒ์ด์ฌ์—์„œ ์ง€์›ํ•˜๋Š” ๋‹ค์–‘ํ•œ ๋ผ์ด๋ธŒ๋Ÿฌ๋ฆฌ์—์„œ๋Š” ๋งŽ์€ ์†์‹คํ•จ์ˆ˜๋ฅผ ์ง€์›ํ•œ๋‹ค. ํ•ด๋‹น ํฌ์ŠคํŒ…์—์„œ๋Š”..

[keras] CNN ๋ชจ๋ธ ๊ตฌ์„ฑ ์‚ดํŽด๋ณด๊ธฐ

๋‚ด๊ฐ€ ํ—ท๊ฐˆ๋ ค์„œ ์ •๋ฆฌ! ๋ชจ๋ธ ํ•˜๋‚˜๋ฅผ ์˜ˆ์‹œ๋กœ ๋“ค์–ด์„œ ๋งค๊ฐœ๋ณ€์ˆ˜๋ฅผ ์„ค๋ช…ํ•ด๋ณด๊ฒ ๋‹ค. 1. ๋ชจ๋ธ ์ƒ์„ฑ model = models.Sequential() keras์˜ Sequential ๋ชจ๋ธ ์‚ฌ์šฉ ๊ณ„์ธต์„ ์„ ํ˜•์œผ๋กœ ์Œ“์€ ๋ชจ๋ธ์„ ์˜๋ฏธ 2. ์ธต (Layer) ์ถ”๊ฐ€ ๋ชจ๋ธ ๊ตฌ์„ฑ INPUT → ์ฝ˜๋ณผ๋ฃจ์…˜ ๋ ˆ์ด์–ด (Conv2D) → ReLU ํ•จ์ˆ˜ → ๋งฅ์Šค ํ’€๋ง ๋ ˆ์ด์–ด(MaxPooling2D) → ์ฝ˜๋ณผ๋ฃจ์…˜ ๋ ˆ์ด์–ด (Conv2D) → ReLU ํ•จ์ˆ˜ → ๋งฅ์Šค ํ’€๋ง ๋ ˆ์ด์–ด(MaxPooling2D) → ์ฝ˜๋ณผ๋ฃจ์…˜ ๋ ˆ์ด์–ด (Conv2D) → ์™„์ „์—ฐ๊ฒฐ๊ณ„์ธต (Dense) → softmaxํ•จ์ˆ˜ → OUTPUT model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28,28,1))) ..

[tensorflow] ํ…์ŠคํŠธ ๋ถ„๋ฅ˜ (Text Classification) - IMDB

์‹œ์ž‘ํ•˜๋ฉฐ TF certification ์„ ์ทจ๋“ํ•˜๊ธฐ ์œ„ํ•ด ํŠœํ† ๋ฆฌ์–ผ๋ถ€ํ„ฐ ๊ณต๋ถ€์ค‘์ด๋‹ค. ์˜ˆ์ƒ๋ณด๋‹ค ์˜ค๋žœ ์‹œ๊ฐ„์ด ๊ฑธ๋ฆด ๊ฒƒ๊ฐ™๋‹ค .. ๊ทธ๋ž˜๋„ ์ตœ๋Œ€ํ•œ ๋‹จ๊ธฐ๊ฐ„์— ์ทจ๋“ํ•˜๋Š”๊ฒƒ์„ ๋ชฉํ‘œ๋กœ ํ•œ๋‹ค!!!!!! ํ™”์ดํŒ… ๋ณธ ํฌ์ŠคํŒ…์—์„œ๋Š” ํ…์ŠคํŠธ ๋ถ„๋ฅ˜๋ฅผ ์œ„ํ•ด ๋‰ด๋Ÿด ๋„คํŠธ์›Œํฌ ํ›ˆ๋ จ์‹œํ‚ค๋Š” ์˜ˆ์ œ๋ฅผ ์‚ดํŽด๋ณธ๋‹ค. ํ…์„œํ”Œ๋กœ์šฐ ๊ณต์‹ ๋ฌธ์„œ ํŠœํ† ๋ฆฌ์–ผ์„ ์ฐธ๊ณ ํ•˜์—ฌ ๋ฆฌ๋ทฐํ•˜๋Š” ํ˜•์‹์˜ ๊ธ€์ด ๋˜๊ฒ ๋‹ค! www.tensorflow.org/tutorials/keras/text_classification ์˜ํ™” ๋ฆฌ๋ทฐ๋ฅผ ์‚ฌ์šฉํ•œ ํ…์ŠคํŠธ ๋ถ„๋ฅ˜ | TensorFlow Core Note: ์ด ๋ฌธ์„œ๋Š” ํ…์„œํ”Œ๋กœ ์ปค๋ฎค๋‹ˆํ‹ฐ์—์„œ ๋ฒˆ์—ญํ–ˆ์Šต๋‹ˆ๋‹ค. ์ปค๋ฎค๋‹ˆํ‹ฐ ๋ฒˆ์—ญ ํ™œ๋™์˜ ํŠน์„ฑ์ƒ ์ •ํ™•ํ•œ ๋ฒˆ์—ญ๊ณผ ์ตœ์‹  ๋‚ด์šฉ์„ ๋ฐ˜์˜ํ•˜๊ธฐ ์œ„ํ•ด ๋…ธ๋ ฅํ•จ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ๊ณต์‹ ์˜๋ฌธ ๋ฌธ์„œ์˜ ๋‚ด์šฉ๊ณผ ์ผ์น˜ํ•˜์ง€ ์•Š์„ ์ˆ˜ www.tensorfl..

[tensorflow] imdb.load_data ์˜ค๋ฅ˜ ํ•ด๊ฒฐ - Object arrays cannot be loaded when allow_pickle=False

didu-story.tistory.com/25?category=903386 [tensorflow] ํ…์ŠคํŠธ ๋ถ„๋ฅ˜ (Text Classification) - IMDB ์‹œ์ž‘ํ•˜๋ฉฐ TF certification ์„ ์ทจ๋“ํ•˜๊ธฐ ์œ„ํ•ด ํŠœํ† ๋ฆฌ์–ผ๋ถ€ํ„ฐ ๊ณต๋ถ€์ค‘์ด๋‹ค. ์˜ˆ์ƒ๋ณด๋‹ค ์˜ค๋žœ ์‹œ๊ฐ„์ด ๊ฑธ๋ฆด ๊ฒƒ๊ฐ™๋‹ค .. ๊ทธ๋ž˜๋„ ์ตœ๋Œ€ํ•œ ๋‹จ๊ธฐ๊ฐ„์— ์ทจ๋“ํ•˜๋Š”๊ฒƒ์„ ๋ชฉํ‘œ๋กœ ํ•œ๋‹ค!!!!!! ํ™”์ดํŒ… ๋ณธ ํฌ์ŠคํŒ…์—์„œ๋Š” ํ…์Šค didu-story.tistory.com ์œ„ ํฌ์ŠคํŒ…์„ ์ง„ํ–‰ํ•˜๋˜ ์ค‘ ๋ฐ์ดํ„ฐ์…‹์„ ๋‹ค์šด๋กœ๋“œ ํ•˜๋Š” ๊ณผ์ •์—์„œ ์•„๋ž˜์™€ ๊ฐ™์€ ์˜ค๋ฅ˜๊ฐ€ ๋ฐœ์ƒํ–ˆ๋‹ค. from keras.datasets import imdb (train_data, train_labels), (test_data, test_labels) = imdb.load_data(num_words=10000..

[๋จธ์‹ ๋Ÿฌ๋‹] ์ง€๋„ํ•™์Šต / ๋น„์ง€๋„ํ•™์Šต ํŠน์ง•๊ณผ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์˜ˆ์‹œ

์ง€๋„ํ•™์Šต (Supervised Learning ์•Œ๊ณ ๋ฆฌ์ฆ˜์— ์ฃผ์ž…๋˜๋Š” ํ›ˆ๋ จ๋ฐ์ดํ„ฐ์— ๋ ˆ์ด๋ธ”(y๊ฐ’)์ด ํฌํ•จ๋˜์–ด ์žˆ๋‹ค. ์˜ˆ) ๊ณ ์–‘์ด์‚ฌ์ง„(1,0), ๊ฐ•์•„์ง€์‚ฌ์ง„ (0,1) ์ด๋Ÿฐ์‹์œผ๋กœ ์ •ํ™•ํ•œ ๋ผ๋ฒจ๋ง์ด ๋˜์–ด์žˆ๋Š” ๋ฐ์ดํ„ฐ๋ฅผ ํ›ˆ๋ จ๋ฐ์ดํ„ฐ๋กœ ์ด์šฉํ•œ๋‹ค. ๋Œ€ํ‘œ์ ์ธ ์ง€๋„ํ•™์Šต ๋ฐฉ๋ฒ• ๋ถ„๋ฅ˜ ์ข…๋ฅ˜๋ฅผ ์˜ˆ์ธกํ•˜๋Š” ๊ฒƒ ๋ฐ์ดํ„ฐ๋ฅผ ํŠน์ • ๋ผ๋ฒจ๊ฐ’(y๊ฐ’)์œผ๋กœ ๋ถ„๋ฅ˜(์˜ˆ์ธก)ํ•˜๋Š” ์ž‘์—… ์ŠคํŒธ๋ถ„๋ฅ˜ ( ์ŠคํŒธ์ด๋‹ค (1) / ์ŠคํŒธ ์•„๋‹ˆ๋‹ค (0) ) : ๋ณดํ†ต Yes / No ๋ฅผ 1๊ณผ 0์œผ๋กœ ํ‘œํ˜„ํ•œ๋‹ค. - ์ด์ง„๋ถ„๋ฅ˜ ์–ด๋–ค ๋ฐ์ดํ„ฐ์— ๋Œ€ํ•ด ์—ฌ๋Ÿฌ๊ฐ’ ์ค‘ ํ•˜๋‚˜๋กœ ๋ถ„๋ฅ˜ํ•˜๊ธฐ๋„ ํ•œ๋‹ค. - ๋‹ค์ค‘๋ถ„๋ฅ˜ ๊ณ ์–‘์ด (1,0,0) / ๊ฐ•์•„์ง€ (0,1,0) / ํ† ๋ผ (0,0,1)... ํšŒ๊ท€ ์—ฐ์†๋œ ๊ฐ’์„ ์˜ˆ์ธกํ•˜๋Š” ๊ฒƒ ์–ด๋–ค ๋ฐ์ดํ„ฐ๋“ค์˜ ํŠน์ง• (feature)์„ ํ† ๋Œ€๋กœ ๊ฐ’(์ˆ˜์น˜)์„ ์˜ˆ์ธก ํ•˜๋Š” ์ž‘์—… ์˜ˆ) feat..

๋ฐ˜์‘ํ˜•