Covers the end to end preparation of raw data for analysis and modeling in machine learning and artificial intelligence. Topics include data collection and ingestion, data quality assessment, detecting and handling missing values with deletion or various imputation strategies, identifying and treating outliers, removing duplicates, and standardizing formats such as dates and categorical labels. Includes data type conversions, categorical variable encoding, feature scaling and normalization, standardization to zero mean and unit variance, and guidance on when each is appropriate given model choice. Covers feature engineering and selection, addressing class imbalance with sampling and weighting methods, and domain specific preprocessing such as data augmentation for computer vision and text preprocessing for natural language processing. Emphasizes correct order of operations, reproducible pipelines, splitting data into training validation and test sets, cross validation practices, and documenting preprocessing decisions and their impact on model performance. Also explains which models are sensitive to feature scale, common pitfalls, and evaluation strategies to ensure preprocessing does not leak information.
HardTechnical
0 practiced
You have an imbalanced, time-dependent churn dataset where the positive churn rate is low and changes over time. Propose a full approach from preprocessing to evaluation: how to split data to avoid leakage, resampling techniques that respect time order, model choices (e.g., time-to-event/survival vs classification), evaluation metrics tied to business goals (revenue retention, lift), and deployment considerations for updating models over time.
MediumTechnical
0 practiced
Explain why feature scaling is important before applying Principal Component Analysis (PCA). Provide a concise numeric or conceptual example showing how an unscaled dataset can produce misleading principal components and describe when it might be acceptable not to scale before PCA.
EasyTechnical
0 practiced
You have customer records with slightly different spellings and address variants. Describe a practical end-to-end approach to detect and merge near-duplicates: blocking/indexing strategies, choice of string similarity metrics (Levenshtein, token-set ratio), threshold tuning, manual review workflows, libraries or tools you would use (e.g., rapidfuzz, dedupe), and how to measure precision/recall of merges.
HardSystem Design
0 practiced
Propose a system and conventions to version and document preprocessing decisions (imputation methods, encoders, scalers, transformation code) so that analysts, auditors, and ML engineers can reproduce past model runs and dashboards. Specify metadata schema (fields to capture), storage options (Git, artifact store, database), human-readable documentation, and change-management processes including approval and rollout.
EasyBehavioral
0 practiced
Tell me about a time you found a data quality issue that affected a business report or dashboard. Describe the Situation, Task, Actions you took to diagnose and fix the problem, how you communicated with stakeholders, and the measurable Result. If you don't have an example, outline how you would approach such a situation end-to-end.
Unlock Full Question Bank
Get access to hundreds of Data Preprocessing and Handling for AI interview questions and detailed answers.
