1. What does "overfitting" in a machine learning model mean?
- The model performs well on new data
- The model performs poorly on training data
- The model memorizes training data but fails on new data
- The model ignores training data
2. Which metric is most appropriate for evaluating a classification model?
- Mean Squared Error
- Accuracy
- R-squared
- Log Loss
3. Understanding model types is key. What type of machine learning is used when the agent interacts with an environment to learn?
- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning
- Deep Learning
4. What is multi-label classification
- Predicting one label per instance
- Predicting multiple labels per instance
- Predicting numeric values
- Predicting time series
5. What is the purpose of feature scaling?
- To remove irrelevant features
- To normalize feature ranges
- To increase model complexity
- To reduce dataset size
What does one-hot encoding do?
- Compresses numeric features
- Converts categorical variables into binary vectors
- Removes missing values
- Normalizes continuous data
7. What is the goal of Principal Component Analysis (PCA)?
- To increase feature count
- To reduce feature redundancy
- To improve accuracy
- To encode labels
8. What is gradient descent used for in machine learning?
- To increase model accuracy
- To encode features
- To optimize loss functions
- To split datasets
9. Which algorithm is considered a non-parametric method?
- Decision Tree
- Linear Regression
- Logistic Regression
- Naive Bayes
10. What is the main idea behind bagging?
- Combining weak learners sequentially
- Encoding categorical variables
- Reducing dimensionality
- Using multiple models trained on random subsets
11. What is model serialization?
- Training a model
- Saving a model for reuse
- Evaluating a model
- Scaling a model
12. What is data leakage in machine learning?
- Loss of data during training
- Overfitting due to noise
- Using future information during training
- Data corruption
13. What is algorithmic bias?
- Random errors in predictions
- Hardware limitations
- Lack of training data
- Systematic errors due to flawed assumptions
14. What is SHAP used for in AI?
- Data cleaning
- Model training
- Explaining model predictions
- Feature scaling
15. Why is explainability important in AI systems?
- It improves training speed
- It helps users trust decisions
- It reduces model size
- It increases accuracy
1 Comments