NEW
Font size
WorksheetsMachine Learning Fundamentals Quiz
Total questions: 10
Worksheet time: 10mins
How does Machine Learning (ML) differ from the broader field of Artificial Intelligence (AI)?
ML is focused on finding patterns and drawing insights from data, which is actually the definition of Data Science.
ML is a technique used to enable computers to perform general human-like tasks and simulate human behavior.
ML is a subset of AI that focuses specifically on solving problems and making predictions using certain data.
ML is the process of manually programming rules for the machine to follow, which is characteristic of traditional AI.
If you are using a dataset where images of animals are already labeled as 'cat,' 'dog,' or 'lizard,' which type of learning model is most appropriate for training?
Unsupervised Learning
Reinforcement Learning
Semi-supervised Learning
Supervised Learning
What is the primary goal of an Unsupervised Learning algorithm, such as K-Means Clustering, when given a set of unlabeled input data points?
To classify the data into predefined categories using the provided labels.
To predict a future numerical value based on historical trends.
To find patterns, structure, or groupings within the unlabeled data.
To determine the most efficient sequence of actions to maximize a reward.
Which of the following is an example of a qualitative (categorical) feature, which means it has a finite number of categories or groups?
The price of a house, which can be any positive value.
The number of children in a household, which must be a whole number.
A satisfaction rating of 'Bad,' 'Good,' or 'Great.'
The length of a patient's stay in a hospital, measured in days and hours.
The video mentions that quantitative data can be discrete. What does discrete quantitative data typically consist of?
Values that can take on any floating-point number within a given range.
Numerical values that are typically represented by counts or whole numbers (integers).
Data that is divided into a finite number of named categories, like colors or types.
Values that are normalized to be between 0 and 1 before being used by the model.
The process of splitting your data into training, validation, and test sets is essential because using the entire dataset for training can lead to what specific problem when the model encounters new data?
Underfitting, resulting in low performance on both the training and test data.
Loss, which is a metric calculated during training to quantify error.
Overfitting, where the model performs well on the training data but poorly on unseen data.
Slower training times due to the increased size of the training set.
What does the term loss represent in the context of training a machine learning model?
A value quantifying the difference between the model's prediction and the true target value.
The percentage of correct predictions the model makes on a given dataset.
The speed at which the model adjusts its parameters during the training process.
A technique used to select the most important features from the input data.
In the code example, the Standard Scalar from sklearn.preprocessing is imported. What is the main function of this tool in data preparation?
To convert qualitative (categorical) features into numerical values (e.g., using one-hot encoding).
To standardize the data by centering it around zero and scaling it to unit variance.
To find the best combination of features to reduce the dimensionality of the dataset.
To impute, or fill in, missing values within the columns of the dataset.
What is the main objective of using Principal Component Analysis (PCA), as demonstrated in the example that reduced a 7-dimensional dataset to 2 dimensions?
To cluster data points into groups based on their similarity.
To reduce the number of features (dimensions) while retaining the maximum amount of original information/variance.
To assign a probability to each data point belonging to a certain class.
To scale all numerical features so they have a minimum value of 0 and a maximum value of 1.
You train a model and find it has a 99% accuracy on the training data but only a 65% accuracy on the test data. What is the most likely issue?
The model is underfitting, meaning it is too simple to capture the complexity of the data.
The model has achieved perfect generalization and is ready for deployment.
The data is perfectly balanced, so no issues with class weights exist.
The model is overfitting, having memorized the training data and failing to generalize to new examples.
