record.md

3 class problem

kmeans++

Converged at iteration 25 Accuracy: 0.9523809523809523

softkmeans

Converged at iteration 30 Accuracy: 0.9047619047619048

mlp

Epoch 1/1000, Train Loss: 25.9276, Test Loss: 25.5430 Epoch 40/1000, Train Loss: 0.7140, Test Loss: 0.6661 Epoch 80/1000, Train Loss: 0.4441, Test Loss: 0.4046 Epoch 120/1000, Train Loss: 0.3791, Test Loss: 0.3347 Epoch 160/1000, Train Loss: 0.3408, Test Loss: 0.3162 Epoch 200/1000, Train Loss: 0.3044, Test Loss: 0.2764 Epoch 240/1000, Train Loss: 0.2856, Test Loss: 0.2622 Epoch 280/1000, Train Loss: 0.2730, Test Loss: 0.2635 Epoch 320/1000, Train Loss: 0.2612, Test Loss: 0.2430 Epoch 360/1000, Train Loss: 0.2565, Test Loss: 0.2318 Epoch 400/1000, Train Loss: 0.2470, Test Loss: 0.2285 Epoch 440/1000, Train Loss: 0.2586, Test Loss: 0.2359 Epoch 480/1000, Train Loss: 0.2367, Test Loss: 0.2371 Epoch 520/1000, Train Loss: 0.2333, Test Loss: 0.2216 Epoch 560/1000, Train Loss: 0.2216, Test Loss: 0.2187 Epoch 600/1000, Train Loss: 0.2195, Test Loss: 0.2276 Epoch 640/1000, Train Loss: 0.2126, Test Loss: 0.2236 Epoch 680/1000, Train Loss: 0.2077, Test Loss: 0.2207 Epoch 720/1000, Train Loss: 0.2029, Test Loss: 0.2176 Epoch 760/1000, Train Loss: 0.1994, Test Loss: 0.2170 Epoch 800/1000, Train Loss: 0.1946, Test Loss: 0.2154 Epoch 840/1000, Train Loss: 0.1921, Test Loss: 0.2146 Epoch 880/1000, Train Loss: 0.1890, Test Loss: 0.2202 Epoch 920/1000, Train Loss: 0.1840, Test Loss: 0.2092 Epoch 960/1000, Train Loss: 0.1823, Test Loss: 0.2157 Epoch 1000/1000, Train Loss: 0.1776, Test Loss: 0.2115 Accuracy: 0.9523809523809523

pca, nonlinear autoencoder and kmeans++, softkmeans

pca and kmeans++

Converged at iteration 40 Accuracy: 0.9285714285714286

pca and softkmeans

Converged at iteration 38 Accuracy: 0.9285714285714286

nonlinear autoencoder and kmeans++

Converged at iteration 35 Accuracy: 0.9523809523809523

nonlinear autoencoder and softkmeans

Converged at iteration 52 Accuracy: 0.9285714285714286

svm

linear kernal

Early stopping at iteration 6985 Early stopping at iteration 11750 Early stopping at iteration 18295 Accuracy: 0.9047619047619048

gaussian kernal

Early stopping at iteration 2496 Early stopping at iteration 2590 Early stopping at iteration 3102 Accuracy: 0.9523809523809523

adaptive boosting

Accuracy: 0.9285714285714286

binary problem

Removing class: 3

mlp: 1.0 svm linear: 0.9285714285714286 svm gaussian: 0.9642857142857143 adaboost: 0.9642857142857143

Removing class: 2

mlp: 0.9285714285714286 svm linear: 0.9285714285714286 svm gaussian: 0.9285714285714286 adaboost: 0.9642857142857143

Removing class: 1

mlp: 1.0 svm linear: 1.0 svm gaussian: 1.0 adaboost: 1.0