FormatMultiple choice, multiple answerTypeSpecialtyDelivery MethodTesting center or online proctored examTime180 minutes to complete the examCost300 USD (Practice exam: 40 USD)LanguageAvailable in English, Japanese, Korean, and Simplified ChineseThe AWS Certified Machine Learning - Specialty certification is intended for individuals who perform a development or data science role. It validates a candidate’s ability to design, implement, deploy, and maintain machine learning (ML) solutions for given business problems. This course is structured into the four domains tested by this exam: data engineering, exploratory data analysis, modeling, and machine learning implementation and operations. Just some of the topics we’ll cover include: S3 data lakesAWS Glue and Glue ETLKinesis data streams, firehose, and video streamsDynamoDBData Pipelines, AWS Batch, and Step FunctionsUsing scikit learnData science basicsAthena and QuicksightElastic MapReduce (EMR)Apache Spark and MLLibFeature engineering (imputation, outliers, binning, transforms, encoding, and normalization)Ground TruthDeep Learning basicsTuning neural networks and avoiding overfittingAmazon SageMaker, in depthRegularization techniquesEvaluating machine learning models (precision, recall, F1, confusion matrix, etc.)High-level ML services: Comprehend, Translate, Polly, Transcribe, Lex, Rekognition, and moreSecurity best practices with machine learning on AWSAbilities Validated by the CertificationSelect and justify the appropriate ML approach for a given business problemIdentify appropriate AWS services to implement ML solutionsDesign and implement scalable, cost-optimized, reliable, and secure ML solutionsRecommended Knowledge and Experience1-2 years of experience developing, architecting, or running ML/deep learning workloads on the AWS CloudThe ability to express the intuition behind basic ML algorithmsExperience performing basic hyperparameter optimizationExperience with ML and deep learning frameworksThe ability to follow model-training best practicesThe ability to follow deployment and operational best practices
