Overview

Exam DP-100: Designing and Implementing a Data Science Solution on AzureLatest Update on March 2024. Exam questions that have been carefully selected to ensure your success on the first try. 150+ different questions. More questions will be added soon...Skills measuredSet up an Azure Machine Learning workspace (30-35%)Run experiments and train models (25-30%)Optimize and manage models (20-25%)Deploy and consume models (20-25%)Detail SkillsDefine and prepare the development environment (15-20%)Select development environmentassess the deployment environment constraintsanalyze and recommend tools that meet system requirementsselect the development environmentSet up development environmentcreate an Azure data science environmentconfigure data science work environmentsQuantify the business problemdefine technical success metricsquantify risksPrepare data for modeling (25-30%)Transform data into usable datasetsdevelop data structuresdesign a data sampling strategydesign the data preparation flowPerform Exploratory Data Analysis (EDA)review visual analytics data to discover patterns and determine next stepsidentify anomalies, outliers, and other data inconsistenciescreate descriptive statistics for a datasetCleanse and transform dataresolve anomalies, outliers, and other data inconsistenciesstandardize data formatsset the granularity for dataPerform feature engineering (15-20%)Perform feature extractionperform feature extraction algorithms on numerical dataperform feature extraction algorithms on non-numerical datascale featuresPerform feature selectiondefine the optimality criteriaapply feature selection algorithmsDevelop models (40-45%)Select an algorithmic approachdetermine appropriate performance metricsimplement appropriate algorithmsconsider data preparation steps that are specific to the selected algorithmsSplit datasetsdetermine ideal split based on the nature of the datadetermine number of splitsdetermine relative size of splitsensure splits are balancedIdentify data imbalancesresample a dataset to impose balanceadjust performance metric to resolve imbalancesimplement penalizationTrain the modelselect early stopping criteriatune hyper-parametersEvaluate model performancescore models against evaluation metricsimplement cross-validationidentify and address overfittingidentify root cause of performance resultsData Scientist is most demanded skill of this era. Certified Data Scientist get more chance to get hired than non-certified candidate.

Skills

Reviews