CS 7643 DEEP LEARNING EXAM SCRIPT
WITH VERIFIED QUESTIONS AND
ANSWERS
●● Feature Engineering
Answer: The process of cleaning, transforming, and creating useful
features from raw data before training a machine learning model.
●● Imputation
Answer: Filling in missing data using methods such as mean, median,
mode, constant values, or forward/back fill.
●● Encoding
Answer: Converting categorical data into numeric values so machine
learning algorithms can process it.
●● Scaling
Answer: Adjusting feature values to the same scale so no feature
dominates due to larger numbers.
●● Supervised Learning
Answer: Machine learning using labeled data where the correct answers
are known during training.
,●● Regression
Answer: A supervised learning task that predicts continuous numeric
values.
●● Classification
Answer: A supervised learning task that predicts categories or classes.
●● Training Dataset
Answer: The portion of data used to teach the model patterns (typically
70-80%).
●● Test Dataset
Answer: The portion of data used to evaluate model performance on
unseen data (typically 20-30%).
●● Accuracy
Answer: Number of correct predictions divided by the total number of
predictions.
●● Confusion Matrix
Answer: A table showing True Positives, True Negatives, False
Positives, and False Negatives.
, ●● Feature
Answer: An input variable or attribute used by a model to make
predictions.
●● Target
Answer: The output variable or label the model is trying to predict.
●● Hyperparameter
Answer: A value chosen by the user before training that controls how a
model learns (example: k in KNN).
●● Parameter
Answer: A value learned automatically by the model during training.
●● Manhattan Distance (L1)
Answer: Measures distance by adding the absolute differences between
coordinates (grid-like movement).
●● Euclidean Distance (L2)
Answer: Measures straight-line distance between two points.
●● Unsupervised Learning
WITH VERIFIED QUESTIONS AND
ANSWERS
●● Feature Engineering
Answer: The process of cleaning, transforming, and creating useful
features from raw data before training a machine learning model.
●● Imputation
Answer: Filling in missing data using methods such as mean, median,
mode, constant values, or forward/back fill.
●● Encoding
Answer: Converting categorical data into numeric values so machine
learning algorithms can process it.
●● Scaling
Answer: Adjusting feature values to the same scale so no feature
dominates due to larger numbers.
●● Supervised Learning
Answer: Machine learning using labeled data where the correct answers
are known during training.
,●● Regression
Answer: A supervised learning task that predicts continuous numeric
values.
●● Classification
Answer: A supervised learning task that predicts categories or classes.
●● Training Dataset
Answer: The portion of data used to teach the model patterns (typically
70-80%).
●● Test Dataset
Answer: The portion of data used to evaluate model performance on
unseen data (typically 20-30%).
●● Accuracy
Answer: Number of correct predictions divided by the total number of
predictions.
●● Confusion Matrix
Answer: A table showing True Positives, True Negatives, False
Positives, and False Negatives.
, ●● Feature
Answer: An input variable or attribute used by a model to make
predictions.
●● Target
Answer: The output variable or label the model is trying to predict.
●● Hyperparameter
Answer: A value chosen by the user before training that controls how a
model learns (example: k in KNN).
●● Parameter
Answer: A value learned automatically by the model during training.
●● Manhattan Distance (L1)
Answer: Measures distance by adding the absolute differences between
coordinates (grid-like movement).
●● Euclidean Distance (L2)
Answer: Measures straight-line distance between two points.
●● Unsupervised Learning
