Data processing NABTEB exam questions and answers

Data processing NABTEB exam questions and answers

Here they are;

Question: What is the primary purpose of data preprocessing in machine learning?

a) Enhancing model interpretability

b) Reducing dimensionality

c) Improving data quality

d) Accelerating model training

Answer: c) Improving data quality

Question: Which technique is commonly used for handling missing data in a dataset?

a) Mean imputation

b) Mode imputation

c) Median imputation

d) Random imputation

Answer: a) Mean imputation

Question: Why is scaling important in data preprocessing?

a) To make data visually appealing

b) To speed up data loading

c) To standardize variable ranges

d) To reduce data dimensionality

Answer: c) To standardize variable ranges

Question: What is the purpose of outlier detection in data preprocessing?

a) To remove irrelevant data

b) To identify and handle extreme values

c) To increase dataset size

d) To introduce noise for model robustness

Answer: b) To identify and handle extreme values

Question: Which normalization technique is suitable for handling skewed data?

a) Min-Max scaling

b) Z-score normalization

c) Log transformation

d) Standard deviation scaling

Answer: c) Log transformation

Question: In feature engineering, what does one-hot encoding accomplish?

a) Reducing feature dimensionality

b) Handling missing data

c) Converting categorical variables into binary vectors

d) Scaling numerical features

Answer: c) Converting categorical variables into binary vectors

Question: What is the purpose of cross-validation in the context of data processing?

a) Enhancing model interpretability

b) Evaluating model performance on multiple subsets of data

c) Imputing missing values

d) Increasing the number of features

Answer: b) Evaluating model performance on multiple subsets of data

Question: How does PCA (Principal Component Analysis) contribute to data preprocessing?

a) It handles missing data

b) It reduces dimensionality

c) It normalizes data distribution

d) It transforms categorical variables

Answer: b) It reduces dimensionality

Question: Which technique is suitable for handling categorical variables with ordinal relationships?

a) Label encoding

b) One-hot encoding

c) Binary encoding

d) Target encoding

Answer: a) Label encoding

Question: What is the purpose of data discretization?

a) Handling missing data

b) Converting numerical variables into categorical variables

c) Scaling data

d) Reducing dimensionality

Answer: b) Converting numerical variables into categorical variables

Question: How does the Bag-of-Words representation contribute to natural language processing?

a) It encodes word order

b) It represents text as a set of unordered words

c) It handles grammatical errors

d) It increases sentence complexity

Answer: b) It represents text as a set of unordered words

Question: What role does feature scaling play in the k-nearest neighbors (KNN) algorithm?

a) It improves model interpretability

b) It speeds up the training process

c) It ensures equal contribution of features in distance calculations

d) It reduces the number of neighbors considered

Answer: c) It ensures equal contribution of features in distance calculations

Question: In time series data, what is the purpose of lag features?

a) They handle missing data

b) They encode categorical information

c) They capture temporal dependencies

d) They standardize variable ranges

Answer: c) They capture temporal dependencies

Question: Why might you use feature scaling before applying the k-means clustering algorithm?

a) To visualize data better

b) To handle missing values

c) To standardize variable ranges

d) To increase the number of clusters

Answer: c) To standardize variable ranges

Question: What does the term "data augmentation" refer to in the context of image processing?

a) Increasing the size of the dataset by duplicating records

b) Enhancing model interpretability

c) Generating new training samples by applying transformations to existing data

d) Reducing the number of features

Answer: c) Generating new training samples by applying transformations to existing data

Question: Which technique is used for handling the class imbalance problem in classification tasks?

a) Feature scaling

b) Data discretization

c) Oversampling minority class

d) Principal Component Analysis (PCA)

Answer: c) Oversampling minority class

Question: What is the purpose of stratified sampling in the context of data splitting?

a) It ensures an equal distribution of classes in both training and testing sets

b) It removes outliers from the dataset

c) It handles missing values

d) It increases the training set size

Answer: a) It ensures an equal distribution of classes in both training and testing sets

Question: How does the concept of feature engineering contribute to machine learning models?

a) It improves data visualization

b) It enhances model interpretability

c) It transforms raw data into informative features for model training

d) It reduces the need for data preprocessing

Answer: c) It transforms raw data into informative features for model training

Question: What is the purpose of a validation set in the model training process?

a) To train the model

b) To fine-tune hyperparameters and prevent overfitting

c) To test the model on unseen data

d) To handle missing values

Answer: b) To fine-tune hyperparameters and prevent overfitting

Question: In natural language processing, what is lemmatization?

a) Removing stop words from text

b) Converting words to their base or root form

c) Encoding words as numerical vectors

d) Increasing the vocabulary size

Answer: b) Converting words to their base or root form