MLA-C01 sample questions with answers
The other 190 questions are in the practice tool, with answers revealed as you go.
Start 20 questionsData preparation for ML
- A marketing analytics team has raw customer exports with inconsistent date formats and blank cells. They want a point-and-click workspace where analysts can apply hundreds of ready-made cleaning steps and save the sequence as a reusable recipe, without writing any scripts. Which AWS service fits?
- A computer vision startup needs thousands of images annotated with bounding boxes. It wants a managed labeling service that can draw on its own staff, a vendor team, or a public crowd, and that trains a model to auto-label the easy images so people focus on the hard ones. Which service should it use?
- An analyst wants to run ad hoc ANSI SQL over Parquet files stored in an Amazon S3 bucket. She wants to pay only for the data each query scans and does not want to provision or manage a cluster. Which AWS service should she use?
- A media company stores raw video footage whose retrieval frequency is unpredictable, sometimes daily and sometimes not for months. The team does not want to write lifecycle rules or monitor access patterns. Which S3 option best fits?
- A data analyst with limited Spark experience wants to build a join and a filter over two S3 datasets by arranging nodes on a canvas. The tool should produce the job script automatically. Which AWS Glue capability fits?
- A Glue job reads JSON exports where some records omit a field and others store the same field as a string instead of a number. The team wants a structure that tolerates these mismatches and can later be converted to a Spark DataFrame. Which Glue construct fits?
- A Firehose stream writes clickstream events to S3, and consumers complain about millions of tiny objects. Which setting should be tuned so that records accumulate before each delivery?
- A warehouse stores a very large sales fact table that is repeatedly joined to smaller dimension tables on customer_id. Which distribution style reduces network traffic during those joins?
- A forecasting team builds features from an event log and wants a seven-day trailing average plus a 30-day rolling maximum for each entity. Which operation produces these?
ML model development
- A travel app wants its assistant to check live seat availability and then book a reservation by calling the airline's booking API, with the model deciding when each call is needed. Which Amazon Bedrock feature supports this?
- A data scientist training a convolutional classifier notices the model memorizes its training images. Which regularization method should be added so that, during each training step, a random subset of hidden units is silenced?
- A research lab must train a multi-billion-parameter language model across hundreds of accelerator instances for several weeks. Hardware failures must be detected and faulty nodes replaced automatically without restarting the whole run from scratch. Which SageMaker option fits?
- A pharmaceutical company has millions of unlabeled internal research articles. It wants the foundation model to absorb this specialized vocabulary and writing patterns before any task-specific training, without supplying question-answer pairs. Which customization method does this?
- A courier company predicts delivery durations in minutes. Observed labels stay between 0 and 100 and never sit near zero, and planners want an error figure they can quote as a percentage of the true duration. Which metric should the team report?
- A vision team finds that early epochs on a mixed-difficulty dataset are unstable. They reorder training so clean, obvious images appear first and ambiguous, noisy images are phased in later, and convergence improves. Which training strategy is this?
- While reviewing a convolutional detector, an engineer explains that a neuron in a deep layer reacts to a wide area of the original image. What is this area called?
- Training a deep recurrent network occasionally produces enormous gradient values that wreck an epoch. Which remedy directly addresses this?
- A plain recurrent network fails to learn dependencies that span dozens of time steps. Which explanation and remedy are correct?
Deployment and orchestration of ML workflows
- A team wants a repeatable ML workflow that preprocesses data, trains a model, evaluates it, and registers the model only if its metrics pass a threshold. The whole flow should be defined as a directed graph and integrate natively with the model catalog. Which service is best?
- A team wants to send a small first slice of production traffic to a retrained recommendation model, watch error-rate and latency alarms during a bake period, and then shift all remaining traffic at once if everything looks fine. Which SageMaker traffic shifting mode is described?
- An inference system must run a feature-normalization container followed by a classifier container, so each request passes through both in order behind one endpoint. Which SageMaker deployment option matches this?
- A trading platform needs model inference to survive the loss of an entire AWS Region, with automatic failover and no manual runbook. Which architecture should be used?