[feat] implement 'svm_loss_vectorized' function #1

cs231n_2022/assignment1/cs231n/classifiers/linear_svm.py

@@ -84,7 +84,15 @@ def svm_loss_vectorized(W, X, y, reg):
     #############################################################################
     # *****START OF YOUR CODE (DO NOT DELETE/MODIFY THIS LINE)*****
 
-    pass
+    num_train = X.shape[0]
+
+    scores = np.matmul(X, W)
+    correct_class_scores = scores[range(num_train), y].reshape(-1, 1)
+
+    margins = np.maximum(0, scores - correct_class_scores + 1)
+    margins[range(num_train), y] = 0
+
+    loss = np.sum(margins) / num_train + reg * np.sum(W * W)
 
     # *****END OF YOUR CODE (DO NOT DELETE/MODIFY THIS LINE)*****
 
@@ -98,8 +106,13 @@ def svm_loss_vectorized(W, X, y, reg):
     # loss.                                                                     #
     #############################################################################
     # *****START OF YOUR CODE (DO NOT DELETE/MODIFY THIS LINE)*****
+
+    # ref: https://mainpower4309.tistory.com/28
+    dScores = np.zeros(scores.shape)
+    dScores[margins > 0] = 1
+    dScores[range(num_train), y] -= np.sum(dScores, axis=1)
 
-    pass
+    dW = np.matmul(X.T, dScores) / num_train + 2 * reg * W
 
     # *****END OF YOUR CODE (DO NOT DELETE/MODIFY THIS LINE)*****