Wàcci degrade stokastik ak laaj ak tontu momentum

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

1. Lan la momentum bi di dajale ci diiru tàggat yaram?
   • Limu jamonoy tàggat yaram yépp
   • ab daw (pondéré exponentiel) moyenne ci degrade yi weesu
   • Dayo done yi
   • Taarixu njubte model bi
2. Ci ravine bu gudd te sew, ban jafe-jafe la SGD bu leer di dundu te momentum bi dafay jàppale ci saafara?
   • Dafay fees ci lim
   • Dafay taxawal xayma degrade yi
   • Dafay yokk njàng mi ci saasi
   • Dafay zig-zag ci kaw miir yu gaaw yi, di dem ndànk ci dër bi
3. Benn valër buñ miin ngir koeffisientu momentum mooy:
   • 0.9
   • 0.0
   • 0.5
   • 2.0
4. Ban analogie physique lañuy gëna jëfandikoo ngir fësal momentum?
   • Pendul buy balance ba fàww
   • Ressort buy dellu
   • Bal bu diis buy wàcci
   • ndox muy gelé
5. Lu tollu ci ñaata la momentum bi di yokk jéego bu am solo bi ci yoon wu méngoo sudee beta = 0.9?
   • Lu tollu ci 2 yoon
   • Lu tollu ci 0.1 yoon
   • Amul benn yokk
   • Lu tollu ci 10 yoon
6. Ban optimisatër bu bees moo boole ci xayma saa yu njëkk yu melni momentum?
   • Xeetu Newton
   • Aadama
   • K-dafay tekki
   • Seetug aleet
7. In the update v = beta times v plus gradient, what happens when beta is zero?
   • The old velocity is doubled
   • Every gradient becomes zero
   • The previous velocity is discarded and the current gradient determines the step
   • The parameters stop depending on the learning rate
8. Why can repeated gradients with opposing sideways directions produce less zig-zagging under momentum?
   • Their opposing contributions can partially cancel in accumulated velocity
   • All past gradients are stored as separate parameters
   • The loss becomes convex automatically
   • Momentum ignores the current gradient entirely