{"description":"Test your understanding of the deep learning pipeline, model design, training, and evaluation strategies.","questions":[{"answer":"Data Preparation","number":1,"options":["Deployment","Model Design","Data Preparation","Evaluation"],"question":"What is the first stage in a typical deep learning pipeline?"},{"answer":"To enhance dataset variety and prevent overfitting","number":2,"options":["To compress the dataset","To enhance dataset variety and prevent overfitting","To simplify the model architecture","To remove all outliers"],"question":"What is the main purpose of data augmentation?"},{"answer":"Dropout","number":3,"options":["Increasing the batch size","Using more training epochs","Adding more layers","Dropout"],"question":"Which of the following is a method used to prevent overfitting in neural networks?"},{"answer":"Softmax","number":4,"options":["Sigmoid","ReLU","Tanh","Softmax"],"question":"Which activation function is commonly used in the output layer of a multi-class classification model?"},{"answer":"Size of weight updates during optimization","number":5,"options":["Number of hidden layers","Number of training samples","Size of weight updates during optimization","Length of the input sequence"],"question":"What does the learning rate control in neural network training?"},{"answer":"Adam","number":6,"options":["SGD","RMSProp","Adam","Adagrad"],"question":"Which optimizer combines adaptive learning rates and momentum?"},{"answer":"Recall","number":7,"options":["Accuracy","Loss","Recall","Epochs"],"question":"What metric is most appropriate when class imbalance exists in a classification problem?"},{"answer":"Early stopping","number":8,"options":["Early stopping","Cross-entropy loss","Batch normalization","Increasing the learning rate"],"question":"Which technique helps you decide when to stop training a model to avoid overfitting?"},{"answer":"A subset used for validation while others are for training","number":9,"options":["A training pass over the whole dataset","A layer in the network","A subset used for validation while others are for training","A data augmentation method"],"question":"In k-fold cross-validation, what does each fold represent?"},{"answer":"To find the best-performing configuration","number":10,"options":["To reduce training data","To manually adjust weights","To find the best-performing configuration","To convert the model to binary"],"question":"What is the goal of tuning hyperparameters in deep learning?"}],"title":"Deep Learning Visual Demo"}