ML2 Chatgpt L2 Easy-Medium

What does 'dropout' refer to in neural networks?

• A regularization technique that temporarily removes units from the network

• The process of eliminating features from the dataset

• Decreasing the number of layers in a deep neural network

• Dropping out of the training process early to prevent overfitting

In the context of NLP (Natural Language Processing), what does 'tokenization' refer to?

• Converting text into numerical data

• Breaking down text into smaller pieces, such as words or phrases

• Encrypting sensitive information in text

• Categorizing text into predefined categories

Which of the following is a key characteristic of 'Deep Learning'?

• Utilizes a shallow network architecture

• Capable of feature engineering automatically from raw data

• Does not require large amounts of data to train

• Performs poorly on unstructured data

What role does 'hyperparameter tuning' play in the development of machine learning models?

• It involves adjusting the dataset size to fit the model better.

• It refers to the process of selecting the type of model to use.

• It includes changing the architecture of neural networks to improve performance.

• It entails optimizing the model settings to achieve the best performance.

What is 'feature importance' used to determine in machine learning models?

• The number of features required for the model to operate

• The features that have the most impact on the model's predictions

• The correlation between features and labels

• The features with the highest variance

Which of the following best describes 'ensemble learning'?

• A technique that involves training only one model to perform a task

• A process of combining several models to solve a particular problem

• A strategy to divide the dataset into smaller subsets for individual models

• The use of multiple datasets to improve the performance of a single model

In machine learning, what does 'pruning' refer to?

• Adding more branches to a decision tree

• Removing unnecessary features from the dataset

• Reducing the size of a neural network

• Trimming a decision tree to prevent overfitting

Which algorithm is particularly good at handling large datasets with thousands of features?

• Naïve Bayes

• Decision Trees

• Linear Regression

• K-Means Clustering

How does 'bagging' help improve the performance of machine learning models?

• By aggregating the predictions of multiple models to reduce variance

• By boosting the performance of weak learners

• By increasing the bias in the model

• By reducing the training time significantly

In the context of Support Vector Machines (SVMs), what is a 'kernel'?

• A function used to map the original dataset into a higher-dimensional space

• A method for selecting support vectors

• A technique to reduce the dimensionality of the feature space

• The algorithm used for classification