๋ณธ๋ฌธ ๋ฐ”๋กœ๊ฐ€๊ธฐ

์‹œํ€€์Šค1

[๋”ฅ๋Ÿฌ๋‹] Windowed Dataset ๊ตฌ์„ฑ ์˜ต์…˜ ( Tensorflow ์‹œ๊ณ„์—ด ๋ฐ์ดํ„ฐ ๋ถ„์„) / ๋”ฅ๋Ÿฌ๋‹ ๋ฏธ๋ž˜ ์˜ˆ์ธกํ•˜๊ธฐ Windowed Dataset Tensorflow ์‹œ๊ณ„์—ด ๋ฐ์ดํ„ฐ ๋ถ„์„ window size (์œˆ๋„์šฐ ๊ตฌ์„ฑ) [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 ] ๋ผ๋Š” ๋ฐ์ดํ„ฐ๋“ค์ด ์žˆ์„ ๋•Œ ์ฒซ ๋ฒˆ์งธ๋ถ€ํ„ฐ ๋‹ค์„ฏ๋ฒˆ์งธ์˜ ๋ฐ์ดํ„ฐ, ์ฆ‰ 0, 1, 2, 3, 4๋ฅผ ๊ฐ€์ง€๊ณ  6๋ฒˆ์งธ ๋ฐ์ดํ„ฐ 5์˜ ๊ฐ’์„ ์˜ˆ์ธกํ•  ์ˆ˜ ์žˆ๋‹ค. ์ด ๋•Œ window_size = 5๋ผ๊ณ  ํ•œ๋‹ค. shift ์œ„์—์„œ shift = 1 ์ด๋ผ๋Š” ์˜ต์…˜์„ ์ฃผ๋ฉด [ 0, 1, 2, 3, 4 ] -> 5 [ 1, 2, 3, 4, 5 ] -> 6 [ 2, 3, 4, 5, 6 ] -> 7 [ 3, 4, 5, 6, 7 ] -> 8 [ 4, 5, 6, 7, 8 ] -> 9 ์ด๋Ÿฐ์‹์œผ๋กœ ์˜ˆ์ธกํ•˜๊ฒŒ ๋œ๋‹ค. shift = 2 ๋ผ๋ฉด, [ 0, 1, 2, 3, 4 ] -> 5 [ 2,.. 2022. 2. 22.
๋ฐ˜์‘ํ˜•

ํ‹ฐ์Šคํ† ๋ฆฌํˆด๋ฐ”