Kuiz Kecerunan Lenyap dan Meletup

Jawapan dan penjelasan muncul semasa anda mengambil kuiz.

1. Apakah operasi matematik dalam perambatan belakang yang menjadi punca kecerunan lenyap dan meletup?
   • Mendarab berulang kali faktor setiap lapisan melalui peraturan rantai
   • Menambah kecerunan merentas satu lapisan
   • Membahagikan kerugian dengan saiz kelompok
   • Mengambil punca kuasa dua pemberat
2. Why can sigmoid or tanh units contribute to vanishing gradients when they are saturated?
   • Their derivatives are always larger than one
   • Their derivatives become small in saturated regions, so repeated chain-rule products can shrink
   • They remove the chain rule from backpropagation
   • They force every weight to zero
3. Seni bina manakah yang PALING teruk terjejas oleh masalah kecerunan pada jujukan yang panjang?
   • Perceptron dua lapisan
   • Regresi linear tunggal
   • Rangkaian saraf berulang dibuka dalam beberapa langkah masa
   • Jadual carian
4. Melihat kehilangan latihan tiba-tiba bertukar menjadi NaN kemungkinan besar menunjukkan masalah yang mana?
   • Kecerunan yang hilang
   • Penumpuan sempurna
   • Kadar pembelajaran yang terlalu kecil
   • Kecerunan yang meletup
5. Bagaimanakah sambungan baki (langkau) membantu dengan kecerunan yang hilang?
   • Mereka memberikan kecerunan laluan pintasan terus ke belakang, memintas pendaraban mengecut
   • Mereka mengalih keluar semua bukan linear
   • Mereka meningkatkan kadar pembelajaran secara automatik
   • Mereka memadam lapisan awal
6. In a simplified analysis of a recurrent linear transformation repeated over many steps, what matrix quantity is a rough signal of growth versus decay?
   • The matrix trace relative to the batch size
   • The spectral radius relative to one
   • The number of hidden units alone
   • The determinant relative to the learning rate
7. A backpropagated gradient is multiplied by a factor below one at each of many layers. What trend should you expect?
   • It must become exactly one
   • It grows exponentially regardless of the factor
   • The signal can shrink exponentially with depth
   • It stops depending on earlier layers after one step
8. Why can a recurrent network be especially vulnerable over a long sequence?
   • The network has no hidden state
   • The output layer is used only once
   • The sequence is processed without a chain rule
   • The recurrent transformation is applied repeatedly across timesteps, compounding its gradient effect