3๊ฐ•.md

  • ๋ฐ์ดํ„ฐ ๋ถ„์„ ๋จผ์ €? ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์„ ํƒ ๋จผ์ €?

    • ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์„ ํƒ ๋ณด๋‹ค ๋ฐ์ดํ„ฐ์˜ ํ˜•ํƒœ๋‚˜ ํŠน์ง•์„ ๋จผ์ € ํŒŒ์•…ํ•˜๊ณ  ์ ํ•ฉํ•œ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ฐพ๋Š”๋‹ค.

  • Machine learning == Pattern learning

  • Data๊ฐ€ ๊ฐ€์žฅ ์ค‘์š”!

    • Your neural network is only as good as the data you feed it.

    • Garbage in, Garbage out

  • Data check

    1. Null ๋ฐ์ดํ„ฐ ์ฐพ๊ธฐ

    2. Feature Engineering

    3. Feature Selection

    4. Model Development

    5. Training strategy

    6. Stacknig and Averaging

Last updated