Scribd
Upload
45 views3 pages

Predictive Analytics Answer Key

Uploaded by

Hruthik
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
45 views3 pages

Predictive Analytics Answer Key

Uploaded by

Hruthik
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Predictive Analytics: Answer Key

UNIT I

Question 1:
Supervised learning is a type of machine learning where the model is trained on labeled
data. Example: Predicting house prices using historical data.

Question 2:
Linear regression predicts a continuous outcome based on input features. The least squares
method minimizes the sum of squared errors.

Question 3:
Ridge regression adds an L2 penalty term to reduce overfitting, while Lasso regression adds
an L1 penalty, which helps with feature selection.

Question 4:
Logistic regression predicts binary outcomes (0 or 1) using a logistic function: P(Y=1) = 1 /
(1 + e^(-z)), where z is a linear combination of inputs.

Question 5:
The Perceptron Learning Algorithm is an iterative process for binary classification that
updates weights to find a linear decision boundary.

UNIT II

Question 1:
The bias-variance tradeoff is a balance between model complexity and accuracy. High bias
leads to underfitting, and high variance leads to overfitting.

Question 2:
Cross-validation splits the dataset into multiple subsets to evaluate model performance and
prevent overfitting.

Question 3:
The Bayesian approach incorporates prior information to assess models and calculates
posterior probabilities for model selection.

Question 4:
Effective number of parameters refers to the number of parameters in a model that
significantly influence predictions.
Question 5:
Conditional test error is the error for a fixed training set, while expected test error averages
errors over all possible training datasets.

UNIT III

Question 1:
Generalized Additive Models (GAMs) extend linear models to include non-linear
relationships using smooth functions like splines.

Question 2:
Regression trees predict continuous outcomes by splitting data into subsets, whereas
classification trees predict categorical outcomes.

Question 3:
AdaBoost combines multiple weak learners (e.g., decision trees) to form a strong classifier
by focusing on misclassified samples.

Question 4:
Gradient Boosting minimizes loss iteratively using numerical optimization techniques to
improve predictions.

Question 5:
Decision trees split data based on feature values to form a tree structure for classifying or
predicting outcomes.

UNIT IV

Question 1:
Neural networks consist of input, hidden, and output layers that simulate the human brain
to learn from data.

Question 2:
Backpropagation is an algorithm that trains neural networks by adjusting weights to
minimize the error between predicted and actual outputs.

Question 3:
Support Vector Machines (SVM) find the optimal hyperplane that separates classes with
maximum margin.

Question 4:
The k-Nearest Neighbor (k-NN) method classifies data points based on the majority class of
their closest neighbors.
Question 5:
Kernels in SVMs transform data into higher-dimensional spaces, allowing non-linear
decision boundaries to be identified.

UNIT V

Question 1:
Clustering groups similar data points together. Types include K-Means clustering (partition-
based) and hierarchical clustering.

Question 2:
Principal Component Analysis (PCA) reduces dimensionality by transforming features into a
smaller set of uncorrelated components.

Question 3:
Association rules identify patterns and relationships in data. Example: 'Customers who buy
bread also buy milk.'

Question 4:
Random forests combine multiple decision trees to improve accuracy and reduce overfitting
using bagging techniques.

Question 5:
Unsupervised learning identifies patterns in unlabeled data. Examples include clustering
and dimensionality reduction.

You might also like