A novel criterion for global asymptotic stability of neutral-type neural networks with discrete time delays

Faydasıçok Ö., Arik S.

25th International Conference on Neural Information Processing, ICONIP 2018, Siem-Reab, Kamboçya, 13 - 16 Aralık 2018, ss.353-360, (Tam Metin Bildiri) identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası:
  • Doi Numarası: 10.1007/978-3-030-04179-3_31
  • Basıldığı Şehir: Siem-Reab
  • Basıldığı Ülke: Kamboçya
  • Sayfa Sayıları: ss.353-360
  • İstanbul Üniversitesi-Cerrahpaşa Adresli: Evet

Özet

The main target in this work is to propose a new condition for global asymptotic stability for neutral-type neural networks including constant delay parameters. This newly obtained criterion is derived by making the use a properly modified Lyapunov functional with the employment of Lipschitz activation functions, which establishes a new set of relationships among the constant system parameters of this class of neural networks. The obtained condition is expressed independently of delay parameters and can be easily approved by simply checking the validity of some algebraic equations.