TensorFlow

NumPy

Keras

SciPy

Inferring rules from data and labels

Compiling binary programs

Defining rules in a programming language

Gathering data and labels

The binary compiled code in traditional programming

The rules inferred from the data and labels

The distinct patterns that denote a particular activity

The item created from the data and labels

To measure the loss value

To make guesses using the optimizer

To learn the relationship between input and output data

To specify the number of epochs

a) To calculate the relationship between X and Y

b) To make predictions for unknown X values

c) To evaluate the accuracy of the model

d) To train the model

The model is compiled into a binary program.

The model is used to gather data and labels.

The model is used to make predictions based on the inferred rules.

The model is used to define activity types.

Machine learning infers rules from data and labels, while traditional programming defines rules in a programming language.

Machine learning defines rules in a programming language, while traditional programming infers rules from data and labels.

Machine learning eliminates the need for rules, while traditional programming heavily relies on defined rules.

Machine learning and traditional programming both use predefined rules.

import tensorflow as tf

import numpy as np

import keras

from tensorflow import keras

a) 2

b) 5

c) 7

d) 9

Labels

Rules

Programming code

Models

a) Certainties

b) Probabilities

c) Exact values

d) Randomness

a) To train the model

b) To calculate the relationship between X and Y

c) To make predictions for unknown X values

d) To evaluate the accuracy of the model

The accuracy of the neural network

The number of epochs completed

The size of the input data

How good or bad the neural network is performing

a) TensorFlow.org

b) Google AI-ML course data

c) Compiling layers and neurons

d) Measuring how well the network performs

Machine learning allows for the creation of new scenarios that may not have been possible with traditional programming.

Machine learning enables the compilation of binary programs.

Machine learning simplifies the process of defining rules in a programming language.

Machine learning eliminates the need for data labeling.

a) 0

b) 4

c) 7

d) 9

a) A certainty value

b) A probability value

c) The actual Y value

d) The X value

Inferring rules from data and labels

Compiling binary programs

Defining rules in a programming language

Gathering data and labels

a) The model is not accurate

b) The code is incorrect

c) Neural networks deal with probabilities

d) The data points are insufficient

To specify the number of epochs for training

To measure the loss value for each epoch

To train the neural network by adjusting the weights and biases

To make guesses using the optimizer

a) To define the relationship between two sets of numbers

b) To compile neurons

c) To generate new parameters for new guesses

d) To handle the process of guessing the relationship between the numbers

Prints out the loss value for each epoch

Measures the accuracy of the neural network

Adjusts the weights and biases to improve performance

Specifies the number of epochs for training

a) It improves the accuracy of the model

b) It decreases the accuracy of the model

c) It has no effect on the model's accuracy

d) It makes the model more complex

a) 28x28

b) 30x30

c) 32x32

d) 24x24

Y = 3X + 1

Y = 2X + 1

Y = X + 3

Y = X - 1

a) 2

b) 4

c) 7

d) 9

Yes, all 500 epochs must be completed for accurate results

No, the loss value becomes small enough after 50 epochs

It depends on the size of the input data

The example does not mention the necessity of completing all epochs

a) To define the sequence of layers in the neural network

b) To flatten the input data

c) To add a layer of neurons

d) To apply activation functions to each layer

a) Turns a square into a one-dimensional vector

b) Adds a layer of neurons

c) Applies the relu activation function

d) Picks the biggest value from a set of values

To evaluate the model's performance on unseen data

To train the model on new data

To improve the accuracy of the model

To validate the model's training process

a) Flatten

b) Dense

c) Sequential

d) Softmax

By calling model.evaluate() and passing in the test set

By calling model.predict() and passing in the test set

By calling model.fit() and passing in the test set

By calling model.compile() and passing in the test set

a) To build the model

b) To train the model

c) To optimize the model

d) To evaluate the model

a) Returns the input value

b) Returns 0

c) Returns 1

d) Returns -1

To extract important features from data

To blur inessential features in data

To detect specific patterns in data

All of the above

To reduce the overall amount of information in an image

To increase the size of the image

To detect features in the image

To remove all features from the image

a) Returns [0,0,0,0,1,0,0,0,0]

b) Returns [0.1, 0.1, 0.05, 0.1, 9.5, 0.1, 0.05, 0.05, 0.05]

c) Returns [0.05, 0.05, 0.05, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]

d) Returns [0,0,0,0,0,0,0,0,1]

a) To flatten the input data

b) To add a layer of neurons

c) To turn a square into a one-dimensional vector

d) To pass values of 0 or greater to the next layer in the network

a) To return values between 0 and 1

b) To return values between -1 and 1

c) To return binary values

d) To return the input value

a) By penalizing large weights in the model

b) By increasing the complexity of the model

c) By reducing the number of epochs during training

d) By removing outliers from the dataset

a) To adjust the weights and biases of the model

b) To calculate the accuracy of the model

c) To determine the learning rate of the optimizer

d) To visualize the training process

a) To return values between -1 and 1

b) To return values between 0 and 1

c) To return binary values

d) To return the input value

a) To evaluate the model's performance on unseen data

b) To train the model on new data

c) To improve the accuracy of the model

d) To validate the model's training process

a) Softmax returns values between 0 and 1, while sigmoid returns binary values

b) Softmax returns binary values, while sigmoid returns values between 0 and 1

c) Softmax returns the input value, while sigmoid returns values between -1 and 1

d) Softmax returns the maximum value, while sigmoid returns the minimum value