Q121
Q121 What is the main purpose of a Support Vector Machine (SVM)?
To reduce the number of features
To classify data points
To increase data size
To perform clustering
Q122
Q122 How does SVM handle non-linearly separable data?
By ignoring the data
By using a linear kernel
By applying the kernel trick
By reducing the data size
Q123
Q123 What is the role of the margin in SVM?
To separate the training and test data
To measure the distance between support vectors
To maximize the distance between data points and the hyperplane
To reduce model complexity
Q124
Q124 What is a support vector in SVM?
A data point farthest from the hyperplane
A data point closest to the hyperplane
A misclassified data point
A point used for outlier detection
Q125
Q125 Which function in sklearn is used to implement SVM for classification?
SVC()
SVM()
svm_classifier()
SupportVectorClassifier()
Q126
Q126 How do you specify the kernel type in sklearn’s SVM implementation?
kernel()
svm_kernel()
set_kernel()
kernel_type()
Q127
Q127 How can you tune the regularization parameter (C) in SVM using sklearn?
reg_param()
tune_c()
C
regularization()
Q128
Q128 An SVM model performs poorly on a dataset with overlapping classes. What could improve its performance?
Increase the regularization parameter (C)
Use a linear kernel
Use a non-linear kernel
Reduce the number of support vectors
Q129
Q129 An SVM model is overfitting on the training data. What should be adjusted to mitigate this?
Increase the value of C
Decrease the value of C
Increase the kernel size
Use a larger dataset
Q130
Q130 What is the main advantage of ensemble learning?
Reduces the number of features
Increases model complexity
Improves model performance by combining multiple models
Increases training time
Q131
Q131 What is the key difference between bagging and boosting in ensemble learning?
Bagging focuses on reducing bias
Boosting focuses on reducing variance
Bagging trains models in parallel
Boosting uses a single model
Q132
Q132 What is the role of weak learners in boosting?
To make accurate predictions
To identify outliers
To correct mistakes from previous learners
To increase bias
Q133
Q133 Which function in sklearn is used to implement a Random Forest classifier?
RandomForestClassifier()
ForestClassifier()
RandomForest()
ForestRandom()
Q134
Q134 How do you implement AdaBoost in Python using sklearn?
BoostingAda()
AdaBoostClassifier()
BoostAda()
Adaboost()
Q135
Q135 Which hyperparameter in sklearn's Gradient Boosting Classifier controls the learning rate?
max_depth
learning_rate
n_estimators
boost_size
Q136
Q136 A Random Forest model is overfitting on the training data. What could resolve this?
Increase the number of estimators
Reduce the number of features
Increase the max_depth parameter
Reduce the number of trees
Q137
Q137 A Gradient Boosting model is slow to train and gives diminishing returns. What could improve training speed?
Reduce learning rate
Increase the number of weak learners
Use early stopping
Increase the number of estimators
Q138
Q138 What is the goal of reinforcement learning?
To predict continuous values
To minimize error rate
To maximize cumulative reward
To classify data points
Q139
Q139 What is the role of the agent in reinforcement learning?
To gather data
To provide rewards
To take actions based on policy
To process the environment data
Q140
Q140 What is the trade-off between exploration and exploitation in reinforcement learning?
Exploring only actions that maximize reward
Exploring new actions
Balancing between trying new actions and maximizing known rewards
Ignoring all rewards
Q141
Q141 Which Python library is commonly used to implement reinforcement learning algorithms?
keras
scikit-learn
gym
numpy
Q142
Q142 Which algorithm in reinforcement learning is used to optimize the action-value function?
Q-Learning
Random Forest
Support Vector Machine
Gradient Boosting
Q143
Q143 How do you update the Q-value in the Q-learning algorithm?
By maximizing the reward
By using the Bellman equation
By minimizing error
By using a classifier
Q144
Q144 An agent in a reinforcement learning model is not exploring enough. What could help resolve this?
Decrease the learning rate
Increase the exploration rate (epsilon)
Decrease the reward function
Increase the discount factor
Q145
Q145 A reinforcement learning model is slow to converge. What could speed up the training process?
Use a smaller dataset
Decrease the discount factor
Increase the learning rate
Increase the number of actions
Q146
Q146 What is a primary ethical concern in machine learning?
Data storage
Model accuracy
Bias and fairness
Feature selection
Q147
Q147 What does bias in machine learning models refer to?
The model’s performance
Prejudice in data or decisions
Incorrect features
Inefficient algorithms
Q148
Q148 Why is transparency important in machine learning models?
To improve model accuracy
To ensure data security
To explain model decisions
To reduce model complexity
Q149
Q149 What is the main challenge with algorithmic decision-making in sensitive domains like healthcare and law enforcement?
Improving accuracy
Ensuring fairness and accountability
Reducing data size
Minimizing computational cost
Q150
Q150 How can machine learning practitioners address privacy concerns when handling sensitive data?
Use biased data
Ignore sensitive features
Implement differential privacy
Increase model complexity
