24 Questions
What type of activation functions are used to predict discrete values in classification tasks?
Nonlinear
Which metric is also known as Log Loss?
Logarithmic Loss
What is the purpose of Area Under Curve (AUC) in binary classification?
Performance measurement
What does the True Positive Rate measure in classification models?
Positive predictions that are actually correct
In which direction does data move in a Feedforward neural network architecture?
Forward only
What is the trade-off between recall and precision?
Higher recall means lower precision
What is the goal of the Naïve Bayes algorithm?
To create the best line or decision boundaries
What is the main purpose of the Linear SVM classifier?
To classify datasets into two classes using a single straight line
Which algorithm is primarily utilized in text classification?
Naïve Bayes
What does the term 'Support Vectors' refer to in machine learning?
Extreme points/vectors that help in creating the hyperplane
In machine learning, what is Posterior Probability denoted by P(A|B)?
Probability of hypothesis A on the observed event B
What does the Non-linear SVM classifier help to classify?
Non-linearly separated data
What is the primary use of a Bernoulli classifier?
Document classification problems
What does the topmost node in a decision tree represent?
Feature
What type of data distribution is assumed by a Multinomial naïve Bayes classifier?
Multinomial distribution
In artificial neural networks, what is adjusted during backpropagation?
Weights
What problem is Long Short-Term Memory (LSTM) designed to overcome?
Vanishing gradient
Which neural network architecture is commonly used for image processing?
CNN (Convolutional Neural Network)
What is the primary function of hidden layers in neural networks?
Process inputs by multiplying them by weights, adding them up, and passing them through an activation function
Perceptrons are limited to solving which type of issues?
Linearly separable issues
Forward propagation in neural networks involves which process?
Passing data forward through the network to generate an output
Convolutional Neural Networks (CNNs) introduce non-linearity in models through what?
Activation functions like ReLU or sigmoid
'Weights and connections' are central to the functioning of which early concept of artificial neurons?
Perceptrons (1960s - 1970s)
What concept was introduced in the 1990s that consists of a flowchart-like tree structure with nodes representing features?
Decision tree
Learn about the definitions of P(B|A), P(A), and P(B), as well as different classifiers like Gaussian model, Multinomial Naive Bayes, and Bernoulli classifier used in Bayesian probability. Understand how these concepts are applied in document classification problems.
Make Your Own Quizzes and Flashcards
Convert your notes into interactive study material.
Get started for free