{"description":"Test your knowledge of popular optimization algorithms used to train deep learning models.","questions":[{"answer":"To minimize the loss function during training","number":1,"options":["To visualize the data","To reduce memory usage","To minimize the loss function during training","To increase dataset size"],"question":"What is the main purpose of an optimizer in deep learning?"},{"answer":"Adam","number":2,"options":["SGD","RMSProp","Adam","Adagrad"],"question":"Which optimizer uses both momentum and adaptive learning rates?"},{"answer":"Stochastic Gradient Descent (SGD)","number":3,"options":["Batch Gradient Descent","Stochastic Gradient Descent (SGD)","RMSProp","Adam"],"question":"Which optimizer updates model parameters using a small batch of data?"},{"answer":"It enables faster updates and better generalization","number":4,"options":["It increases memory usage","It slows down convergence","It enables faster updates and better generalization","It removes the need for activation functions"],"question":"Why is using mini-batches beneficial in training deep learning models?"},{"answer":"RMSProp","number":5,"options":["Momentum","Adam","SGD","RMSProp"],"question":"Which optimizer adjusts the learning rate based on a moving average of past squared gradients?"},{"answer":"Accelerate updates in the right direction","number":6,"options":["Take random steps","Avoid computing gradients","Accelerate updates in the right direction","Convert output to binary"],"question":"What does the 'momentum' term in optimization help the model do?"},{"answer":"SGD without momentum","number":7,"options":["Adam","SGD without momentum","RMSProp","Adadelta"],"question":"Which optimizer might suffer from high variance in parameter updates?"},{"answer":"The point where the loss function is lowest overall","number":8,"options":["A randomly selected training point","The highest point on a graph","The point where the loss function is lowest overall","The first parameter initialized"],"question":"What is a global minimum in the context of optimization?"},{"answer":"It combines the benefits of momentum and adaptive learning rates","number":9,"options":["It is the oldest method","It ignores past gradients","It combines the benefits of momentum and adaptive learning rates","It does not require a learning rate"],"question":"Why is Adam widely used in modern deep learning tasks?"},{"answer":"SGD","number":10,"options":["Adam","SGD","RMSProp","Momentum"],"question":"Which optimizer is most likely to struggle with noisy gradients without enhancements?"}],"title":"Optimizers"}