Amazon MLS-C01 Exam Dumps

AWS Certified Machine Learning - Specialty

Amazon MLS-C01 Sample Questions

Question # 1

A Machine Learning Specialist is deciding between building a naive Bayesian model or afull Bayesian networkfor a classification problem. The Specialist computes the Pearson correlation coefficientsbetween eachfeature and finds that their absolute values range between 0.1 to 0.95.Which model describes the underlying data in this situation?

A. A naive Bayesian model, since the features are all conditionally independent. 
B. A full Bayesian network, since the features are all conditionally independent. 
C. A naive Bayesian model, since some of the features are statistically dependent. 
D. A full Bayesian network, since some of the features are statistically dependent. 

Answer: C

Show Answer

Question # 2

A retail company wants to combine its customer orders with the product description datafrom its product catalog. The structure and format of the records in each dataset isdifferent. A data analyst tried to use a spreadsheet to combine the datasets, but the effortresulted in duplicate records and records that were not properly combined. The companyneeds a solution that it can use to combine similar records from the two datasets andremove any duplicates.Which solution will meet these requirements?

A. Use an AWS Lambda function to process the data. Use two arrays to compare equalstrings in the fields from the two datasets and remove any duplicates. 
B. Create AWS Glue crawlers for reading and populating the AWS Glue Data Catalog. Callthe AWS Glue SearchTables API operation to perform a fuzzy-matching search on the twodatasets, and cleanse the data accordingly. 
C. Create AWS Glue crawlers for reading and populating the AWS Glue Data Catalog. Usethe FindMatches transform to cleanse the data. 
D. Create an AWS Lake Formation custom transform. Run a transformation for matchingproducts from the Lake Formation console to cleanse the data automatically. 

Answer: D

Show Answer

Question # 3

A logistics company needs a forecast model to predict next month's inventory requirementsfor a single item in 10 warehouses. A machine learning specialist uses Amazon Forecast todevelop a forecast model from 3 years of monthly data. There is no missing data. Thespecialist selects the DeepAR+ algorithm to train a predictor. The predictor means absolutepercentage error (MAPE) is much larger than the MAPE produced by the current humanforecasters.Which changes to the CreatePredictor API call could improve the MAPE? (Choose two.)

A. Set PerformAutoML to true. 
B. Set ForecastHorizon to 4. 
C. Set ForecastFrequency to W for weekly. 
D. Set PerformHPO to true. 
E. Set FeaturizationMethodName to filling. 

Answer: C,D

Show Answer

Question # 4

A library is developing an automatic book-borrowing system that uses AmazonRekognition. Images of library members’ faces are stored in an Amazon S3 bucket. Whenmembers borrow books, the Amazon Rekognition CompareFaces API operation comparesreal faces against the stored faces in Amazon S3.The library needs to improve security by making sure that images are encrypted at rest.Also, when the images are used with Amazon Rekognition. they need to be encrypted intransit. The library also must ensure that the images are not used to improve AmazonRekognition as a service.How should a machine learning specialist architect the solution to satisfy theserequirements?

A. Enable server-side encryption on the S3 bucket. Submit an AWS Support ticket to optout of allowing images to be used for improving the service, and follow the processprovided by AWS Support. 
B. Switch to using an Amazon Rekognition collection to store the images. Use theIndexFaces and SearchFacesByImage API operations instead of the CompareFaces APIoperation. 
C. Switch to using the AWS GovCloud (US) Region for Amazon S3 to store images and forAmazon Rekognition to compare faces. Set up a VPN connection and only call the AmazonRekognition API operations through the VPN. 
D. Enable client-side encryption on the S3 bucket. Set up a VPN connection and only callthe Amazon Rekognition API operations through the VPN. 

Answer: B

Show Answer

Question # 5

A Data Scientist is building a model to predict customer churn using a dataset of 100continuous numericalfeatures. The Marketing team has not provided any insight about which features arerelevant for churnprediction. The Marketing team wants to interpret the model and see the direct impact ofrelevant features onthe model outcome. While training a logistic regression model, the Data Scientist observesthat there is a widegap between the training and validation set accuracy.Which methods can the Data Scientist use to improve the model performance and satisfythe Marketing team’sneeds? (Choose two.)

A. Add L1 regularization to the classifier 
B. Add features to the dataset 
C. Perform recursive feature elimination 
D. Perform t-distributed stochastic neighbor embedding (t-SNE) 
E. Perform linear discriminant analysis 

Answer: B,E

Show Answer

Question # 6

A bank wants to launch a low-rate credit promotion. The bank is located in a town thatrecently experienced economic hardship. Only some of the bank's customers were affectedby the crisis, so the bank's credit team must identify which customers to target with thepromotion. However, the credit team wants to make sure that loyal customers' full credithistory is considered when the decision is made.The bank's data science team developed a model that classifies account transactions andunderstands credit eligibility. The data science team used the XGBoost algorithm to trainthe model. The team used 7 years of bank transaction historical data for training andhyperparameter tuning over the course of several days.The accuracy of the model is sufficient, but the credit team is struggling to explainaccurately why the model denies credit to some customers. The credit team has almost noskill in data science.What should the data science team do to address this issue in the MOST operationallyefficient manner?

A. Use Amazon SageMaker Studio to rebuild the model. Create a notebook that uses theXGBoost training container to perform model training. Deploy the model at an endpoint.Enable Amazon SageMaker Model Monitor to store inferences. Use the inferences tocreate Shapley values that help explain model behavior. Create a chart that shows featuresand SHapley Additive exPlanations (SHAP) values to explain to the credit team how thefeatures affect the model outcomes. 
B. Use Amazon SageMaker Studio to rebuild the model. Create a notebook that uses theXGBoost training container to perform model training. Activate Amazon SageMakerDebugger, and configure it to calculate and collect Shapley values. Create a chart thatshows features and SHapley Additive exPlanations (SHAP) values to explain to the creditteam how the features affect the model outcomes. 
C. Create an Amazon SageMaker notebook instance. Use the notebook instance and theXGBoost library to locally retrain the model. Use the plot_importance() method in the Python XGBoost interface to create a feature importance chart. Use that chart to explain tothe credit team how the features affect the model outcomes. 
D. Use Amazon SageMaker Studio to rebuild the model. Create a notebook that uses theXGBoost training container to perform model training. Deploy the model at an endpoint.Use Amazon SageMaker Processing to post-analyze the model and create a feature importance explainability chartautomatically for the credit team. 

Answer: C

Show Answer

Question # 7

A company supplies wholesale clothing to thousands of retail stores. A data scientist mustcreate a model that predicts the daily sales volume for each item for each store. The datascientist discovers that more than half of the stores have been in business for less than 6months. Sales data is highly consistent from week to week. Daily data from the databasehas been aggregated weekly, and weeks with no sales are omitted from the currentdataset. Five years (100 MB) of sales data is available in Amazon S3.Which factors will adversely impact the performance of the forecast model to be developed,and which actions should the data scientist take to mitigate them? (Choose two.)

A. Detecting seasonality for the majority of stores will be an issue. Request categoricaldata to relate new stores with similar stores that have more historical data. 
B. The sales data does not have enough variance. Request external sales data from otherindustries to improve the model's ability to generalize. 
C. Sales data is aggregated by week. Request daily sales data from the source databaseto enable building a daily model. 
D. The sales data is missing zero entries for item sales. Request that item sales data fromthe source database include zero entries to enable building the model. 
E. Only 100 MB of sales data is available in Amazon S3. Request 10 years of sales data,which would provide 200 MB of training data for the model. 

Answer: A,B

Show Answer

Question # 8

A credit card company wants to build a credit scoring model to help predict whether a newcredit card applicantwill default on a credit card payment. The company has collected data from a large numberof sources withthousands of raw attributes. Early experiments to train a classification model revealed thatmany attributes arehighly correlated, the large number of features slows down the training speed significantly,and that there aresome overfitting issues.The Data Scientist on this project would like to speed up the model training time withoutlosing a lot ofinformation from the original dataset.Which feature engineering technique should the Data Scientist use to meet the objectives?

A. Run self-correlation on all features and remove highly correlated features 
B. Normalize all numerical values to be between 0 and 1 
C. Use an autoencoder or principal component analysis (PCA) to replace original featureswith new features 
D. Cluster raw data using k-means and use sample data from each cluster to build a newdataset 

Answer: B

Show Answer