15. Which activation function outputs values between 4 and 1? a. Tanh b. ReLU c. Sigmoid d. Softmax 16. Which of the following is a state -of-the-art G GPU/hardware accelerator or for deep learning? a. Raspberry Pi b. Arduino c. H100 d. Intel Core 17 17. What distinguishes AGI from Narrow Al? a. AGl is limited to a single domain b. AGl only works with supervised learning c. AGI can solve a wide range of tasks like humans d. AGI relies solely on rule-based systems 18. What is the Perceptron in the context of neural networks? a. The simplest form of a neural network with a single layer b. A complex neural network with multiple hidden layers c. A type of convolutional layer d. Ignoring basic neural network structures 19. What is the primary purpose of the error backpropagation algorithm in training neural networks? a. To adjust the weights of the network based on prediction errors b. To introduce randomness in neural network training c. Ignoring the training process d. To increase computational complexity 20. What is the primary purpose of an activation function in a neural network? a. To initialize weights b. To introduce non -linearity c. To store training data d. To connect neurons 21. What is the name of the method described below? __ is the process of predicting the category of given data points. Categories are sometimes called as targets or labels. a. Classification b. Clustering c. Dimensionality Reduction d. Feature Extraction 22. Which of the following open-source licenses al allows commercial ial use of the software? a. MIT b. GPL c. Apache d. All of the above