Best AWS Tool to Merge and Clean Datasets for ML?
An ML engineer receives datasets that contain missing values, duplicates, and extreme outliers. The ML engineer must consolidate these datasets into a single data frame and must prepare the data for ML. Which solution will meet these requirements?
Community Insight
The question tests which SageMaker capability handles ingestion plus data preparation (imputation, deduplication, outlier handling) in one place — the trap is confusing it with SageMaker Ground Truth or data labeling, which produce labels rather than clean data.
Amazon SageMaker Data Wrangler is the purpose-built tool for importing multiple datasets, joining them into one data frame, and applying built-in cleansing transforms for missing values, duplicates, and outliers. The page confirms option A as the correct answer and explains why Ground Truth, data labeling, and manual code generation do not satisfy the requirement.
Learners pick Ground Truth or SageMaker data labeling because the phrase "prepare the data" is misread as annotation work; those services add labels for supervised learning and do not remove duplicates, fill missing values, or treat outliers.
Community Discussion (4 comments)
Comments & Corrections
No comments yet — spotted an error or have a note? Share it below.
Expert Analysis
Why the Answer Is Correct
Amazon SageMaker Data Wrangler is explicitly designed for the import-and-prepare workflow described: it can pull data from Amazon S3, Amazon Athena, Amazon Redshift, and Snowflake into a single flow, join multiple datasets into one data frame, and then apply built-in transforms such as imputing missing values, dropping duplicates, and handling outliers. It also provides data-quality and bias reports plus quick model previews, so an ML engineer can consolidate and cleanse data without writing glue code by hand. As GiorgioGss put it, "This is basic use-case for data wrangler". Saransundar's write-up agrees that Data Wrangler "simplifies merging and cleaning datasets", which matches every clause of the requirement.Why the Other Options Are Wrong
Option B fails because Amazon SageMaker Ground Truth is a human-in-the-loop labeling service; it does not consolidate raw datasets into a data frame or perform cleansing transforms. Option D has the same flaw — SageMaker data labeling adds target labels, not imputations or deduplication. Option C is technically possible but relies on manual merging plus generated code snippets, which is slower, harder to reproduce, and does not offer Data Wrangler's built-in outlier and duplicate handling; the requirement asks for a solution, not ad-hoc scripting. Even 78e659d's comment stresses that Data Wrangler "imports multiple sources, joins them into one flow, and ships built-in transforms", the complete requirement in one tool.Community Comment Notes
All four commenters — Saransundar, GiorgioGss, 78e659d, and rinip86277 — converge on option A, so there is no dissent to weigh. Saransundar frames it cleanly by contrasting each distractor: "Ground Truth is for labeling, not cleaning", while 78e659d emphasizes that the whole requirement lands in one tool. rinip86277 simply states "Option A is the best fit", which aligns with the analysis above. The consensus here matches the vendor's documented purpose for Data Wrangler rather than resting on vote count alone.Official Reference
Exam Strategy
Map each verb in the stem to a service capability: "import/consolidate" points to Data Wrangler's data flows and joins, and "missing values, duplicates, outliers" points to its built-in transforms. Whenever a distractor mentions Ground Truth or data labeling, check whether the stem asks for labels or for cleaned features — labeling services never satisfy cleansing requirements.
Frequently Asked Questions
Why can't SageMaker Ground Truth consolidate and clean the datasets?
Ground Truth is a human-in-the-loop labeling service that produces annotations, not a data frame. It offers no built-in imputation, deduplication, or outlier transforms.
Does Data Wrangler handle all three issues: missing values, duplicates, and outliers?
Yes. Data Wrangler ships built-in transforms for imputing missing values, dropping duplicates, and scaling or clipping outliers, all inside one data flow.
Related Analysis
Practice All MLA-C01 Questions
Access 115 questions with complete answers and detailed explanations.
View Full MLA-C01 Practice Test →