Discuss the difference between Euclidean distance and Manhattan distance metrics in the context of KNN.
- Euclidean is faster, Manhattan is more accurate
- Euclidean is for 3D, Manhattan for 2D
- Euclidean is for continuous data, Manhattan for categorical
- Euclidean uses squares, Manhattan uses absolutes
Euclidean distance is the square root of the sum of squared differences, while Manhattan distance is the sum of the absolute differences.
A dataset with very high between-class variance but low within-class variance is given. How would the LDA approach be beneficial here?
- LDA would be the same as PCA
- LDA would perform optimally due to the variance characteristics
- LDA would perform poorly
- LDA would require transformation of the dataset
LDA would "perform optimally" in this scenario, as high between-class variance and low within-class variance align perfectly with its objective of maximizing between-class variance and minimizing within-class variance.
What is the risk of using the same data for both training and testing in a Machine Learning model?
- Increase in accuracy; Reduction in bias
- Increase in complexity; Reduction in training time
- Reduction in training time; Increase in bias
- Risk of overfitting; Unrealistic performance estimates
Using the same data for training and testing leads to the risk of overfitting and provides unrealistic performance estimates. The model will have seen all the data during training, so it might not generalize well to new, unseen instances.
When a Decision Tree is too complex and fits the training data too well, __________ techniques can be applied to simplify the model.
- Bagging
- Boosting
- Normalizing
- Pruning
When a Decision Tree is overfitting (too complex), pruning techniques can be applied to simplify the model. Pruning involves removing branches that have little predictive power, thereby reducing the complexity and the risk of overfitting.
What's the difference between simple linear regression and multiple linear regression?
- One is for classification, the other for regression
- One is linear, the other is nonlinear
- One uses one predictor, the other uses more than one
- One uses time series data, the other doesn't
Simple linear regression involves one independent variable to predict the dependent variable, while multiple linear regression uses two or more independent variables for prediction.
What is the main goal of Artificial Intelligence as a scientific discipline?
- To create complex algorithms
- To improve data storage
- To make money through technology
- To mimic human intelligence and decision-making
The primary goal of Artificial Intelligence is to develop systems that can mimic human intelligence and decision-making processes.
How can Cross-Validation help in hyperparameter tuning?
- By allowing repeated testing on the same validation set
- By improving model accuracy directly
- By providing robust performance estimates to select the best hyperparameters
- By reducing computation time
Cross-Validation enables hyperparameter tuning by providing a robust estimate of the model's performance across different data splits. This process helps to find hyperparameters that generalize well to unseen data, minimizing the risk of overfitting, and allowing a more informed selection of optimal hyperparameters.
How do Ridge and Lasso regularization techniques interact with Polynomial Regression to mitigate overfitting?
- By adding a penalty term to constrain coefficients
- By fitting low-degree polynomials
- By ignoring interaction terms
- By increasing the model's complexity
Ridge and Lasso regularization techniques mitigate overfitting in Polynomial Regression by adding a penalty term to the loss function. This constrains the coefficients, reducing the complexity of the model, and helps in avoiding overfitting.
You need to build a recommendation system with a clear explanation of why certain recommendations are made. How would you approach this considering AI, Machine Learning, and Deep Learning?
- AI, for its rule-based logic
- Deep Learning, for its data-driven insights
- Machine Learning, for its predictive analysis
- nan
Traditional AI models, often based on rule-based logic, may provide clear explanations for recommendations, whereas Machine Learning and Deep Learning models may be less interpretable.
Can you explain how the choice of kernel in SVM affects the decision boundary?
- Changes the color
- Increases computation time
- Reduces data size
- Transforms the feature space
Different kernels transform the feature space in different ways, affecting the shape and position of the decision boundary in SVM.