Pembelajaran Mesin (Machine Learning) Bagian 1
Merupakan catatan pribadi dari pra-kuliah materi perdana (pertemuan kedua)
Basic of Neural Network (NN) :
Basic of Neural Network (NN) :
- Used to predict missing data
- Uses a compilation of data (list) as the reference for prediction
- Sigmoid(x) is used to range between 0 and 1
- Weight starts at random, adapts with each correction
- Square error cost for accuracy is needed : cost (b) = (b-r)^2
- Utilize slope for better accuracy where b = b*lrn*slope
- There are 3 calculation methods for slope: numerical, algebraically, calculus (derivative)
[Q1] Can computer become hyper-intelligent enough to threaten humanity?
Komentar
Posting Komentar
Silakan berkomentar!
Posting Iklan Promosi (kecuali promosi blog) tanpa komentar ke subjek akan dihapus.