{"description":"Test your understanding of image classification, CNN architecture, and the steps in building a digit recognition model.","questions":[{"answer":"Convolutional Neural Networks (CNNs)","number":1,"options":["Recurrent Neural Networks (RNNs)","Feedforward Neural Networks","Convolutional Neural Networks (CNNs)","Radial Basis Function Networks"],"question":"Which type of neural network is most commonly used for image classification tasks?"},{"answer":"To reduce the spatial dimensions of feature maps","number":2,"options":["To perform classification","To normalize pixel values","To reduce the spatial dimensions of feature maps","To apply activation functions"],"question":"What is the primary purpose of a pooling layer in a CNN?"},{"answer":"ReLU","number":3,"options":["Sigmoid","Tanh","ReLU","Softmax"],"question":"Which activation function is most commonly used in convolutional layers?"},{"answer":"Accuracy","number":4,"options":["Mean Squared Error","Cross Entropy","Accuracy","Gradient Magnitude"],"question":"What metric is typically used to evaluate classification accuracy?"},{"answer":"To convert logits to class probabilities","number":5,"options":["To reduce overfitting","To detect edges","To convert logits to class probabilities","To apply max pooling"],"question":"What is the role of the softmax activation function in classification models?"},{"answer":"To ensure numerical stability during training","number":6,"options":["To increase image brightness","To convert images to grayscale","To ensure numerical stability during training","To create labels automatically"],"question":"Why is image normalization important before training a CNN?"},{"answer":"Correct and incorrect predictions by class","number":7,"options":["The total number of features","The architecture of the model","Correct and incorrect predictions by class","The number of epochs completed"],"question":"What does the confusion matrix help you understand about a model?"},{"answer":"Convolutional layer","number":8,"options":["Dense layer","Loss function","Convolutional layer","Dropout layer"],"question":"Which component of a CNN detects features such as edges or textures in an image?"},{"answer":"Visualize a confusion matrix and plot accuracy/loss curves","number":9,"options":["Generate a pie chart","Visualize a confusion matrix and plot accuracy/loss curves","Increase the image size","Convert outputs to grayscale"],"question":"What is typically done after training a CNN to evaluate its performance?"},{"answer":"The model memorizes training data and performs poorly on new data","number":10,"options":["The model performs well on both training and test sets","The model memorizes training data and performs poorly on new data","The model is too shallow","The convolutional filters are too small"],"question":"What does overfitting indicate in a CNN model?"}],"title":"Machine Vision Basics"}