jung-geun/PSO
1
0
Fork 0
mirror of https://github.com/jung-geun/PSO.git synced 2025-12-20 04:50:45 +09:00
Code Issues Packages Projects Releases Wiki Activity

23-05-29 | 2

Browse Source 
처음 초기화를 균일 분포로 랜덤하게 시작함
iris 기준 11 세대만에 99.16 % 에 도달
성능이 매우 높게 나타남
This commit is contained in:
jung-geun
2023-05-29 04:54:20 +09:00
parent 91c6ec965b
commit c5731c6870
5 changed files with 33 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
mnist.py
View File

@@ -73,13 +73,13 @@ x_test, y_test = get_data_test()
# loss = 'poisson'
# loss = 'cosine_similarity'
# loss = 'log_cosh'
# loss = 'huber_loss'
loss = 'huber_loss'
# loss = 'mean_absolute_error'
# loss = 'mean_absolute_percentage_error'
loss = 'mean_squared_error'
# loss = 'mean_squared_error'
pso_mnist = Optimizer(model, loss=loss, n_particles=50, c0=0.4, c1=0.8, w_min=0.75, w_max=1.4)
pso_mnist = Optimizer(model, loss=loss, n_particles=50, c0=0.5, c1=0.8, w_min=0.75, w_max=1.3)
weight, score = pso_mnist.fit(
x_test, y_test, epochs=1000, save=True, save_path="./result/mnist", renewal="acc", empirical_balance=False, Dispersion=True)
pso_mnist.model_save("./result/mnist")