KO‘P VAZNLI BOG‘LANGAN LOGISTIK REGRESSIYA USULI
Keywords:
Kalit so’zlar: Logistik regressiya, neyron, tasniflash, vazn koeffitsientiAbstract
Abstrakt:Ushbu ishda yangi turdagi logistik regressiya modeli taklif qilingan.
Bu modelda vazn koeffitsiyentlari o‘lchami bitta koeffitsiyentdan bir nechta
koeffitsiyentga kengaytiriladi. Ya’ni, har bir kirish va yashirin qatlam o‘rtasida yagona
bog‘lanish koeffitsiyenti emas, balki bir nechta bog‘lanish mavjud bo‘ladi. CIFAR-10,
CDC Diabet va boshqa ma’lumotlar to‘plamlarida o‘tkazilgan hisoblash tajribalari
shuni ko‘rsatdiki, mavjud logistik regressiya modelining ishlashi kirish va yashirin
qatlam o‘rtasidagi bog‘langan vaznlarning o‘lchamini kengaytirish orqali yaxshilanishi
mumkin.
References
Adabiyotlar
1. Brian, K., Ian Q, W., & Teskey, G. C. (2014). An introduction to brain and
behavior (1st Ed.). Ormerod, Spring 2018. ISBN: 10- 1-4641-0601-0.
2. Chen, Y. L., Rinks, D., & Tang, K. (1997). Critical path in an activity network
with time constraints. European Journal of Operational Research, 100(1), 122–
133. https://doi.org/10.1016/S0377-2217(96)00140-3 Cheng, C.,
3. Feng, X., Huang, J., Jiao, Y., & Zhang, S. (2022). ℓ0-Regularized high-
dimensional accelerated failure time model. Computational Statistics & Data
Analysis, 170, 107430.
