PEMANFAATAN PHYSICS INFORMED NEURAL NETWORK MENGGUNAKAN NEURAL NETWORK ALGORITHM UNTUK MENYELESAIKAN PERSAMAAN DIFERENSIAL
Improved performance and convergence speed in neural network training is very important in the development of artificial intelligence applications. The optimizer is the key component responsible for optimizing the weights and biases in a neural network. One of the techniques developed is the Neur...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/75678 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Improved performance and convergence speed in neural network training is very important
in the development of artificial intelligence applications. The optimizer is the key
component responsible for optimizing the weights and biases in a neural network. One
of the techniques developed is the Neural network algorithm optimizer. In recent years
there has been a method used to get an approximation to a solution to partial differential
equations known as the Physics-informed neural network (PINN) method which was
introduced by Raissi in 2019, this method is to model functions that include solutions
to equations differential, using physics information and other physical variables related
to the system being studied. In this study, the Physics-informed neural network (PINN)
method was combined, which generally uses a derivative-based optimizer but in this
study it was combined with the Neural network algorithm optimizer which was first
introduced by Sadollah in 2018 to solve the equation partial differential. |
|---|