К содержимому
learnspaceYOUR NEXT CHAPTER
ПРОСТРАНСТВО ОБУЧЕНИЯ
ГлавнаяКаталог курсовМоё обучениеCoursera

Знания без границ

Учитесь у лучших университетов и компаний мира.

Открыть Coursera
Интеграция
Пространство университета
Моё пространствоСтраница курса
↵
ЯЛичный кабинетСтудент
© 2026 LearnSpaceКаждый день — возможность узнать больше.Помощь
ML Data Pipelines and Communicating AI Insights · LearnSpace
Назад в каталог
courseraАнализ данных

ML Data Pipelines and Communicating AI Insights

Курс от Coursera
Средний≈ 13.9 чАнглийский
О курсеНавыкиПрограммаПреподаватели

О курсе

ML Data Pipelines and Communicating AI Insights focuses on preparing, engineering, and analyzing data to support scalable machine learning systems. In this course, you will learn how to design data pipelines that ingest, process, and validate datasets used for training and evaluating AI models. You will begin by engineering data pipelines that clean, transform, and govern large datasets using modern data processing frameworks. The course then explores techniques for transforming and analyzing data to generate meaningful insights that support machine learning decisions. Next, you will apply exploratory data analysis and feature engineering techniques to improve model performance and evaluate business impact using analytical metrics. You will also learn how to communicate AI insights effectively through visualizations and structured reporting. Finally, the course introduces strategies for breaking down complex machine learning problems into modular components that can be implemented in scalable ML workflows. By the end of this course, you will be able to build reliable data pipelines, perform data-driven analysis, and communicate AI insights that support decision-making. Tools used in this course include Python, Pandas, Apache Spark, PySpark, SQL, and data visualization frameworks.

Навыки, которые вы освоите

Data PipelinesFeature EngineeringData PreprocessingData TransformationModel EvaluationA/B TestingModel TrainingData StorytellingExtract, Transform, LoadData PresentationData-Driven Decision-MakingData QualityData AnalysisPySparkData GovernanceApache SparkMachine LearningData ProcessingData VisualizationPandas (Python Package)

Программа курса

9 модулей · 65 учебных материалов

01Engineer, Validate, and Govern ML Data: Designing ETL Pipelines That Produce ML-Ready Data6 материалов

Designing ETL Pipelines That Produce ML-Ready Data

Welcome and What You'll LearnВидеоGetting Oriented: Your Data Engineering BaselineDIALOGUEWhy ETL Matters for Machine LearningВидеоFoundations of Scalable ETL for MLЧтение

Учитесь у экспертов

Professionals from the Industry

Преподаватель курса

ML Data Pipelines and Communicating AI Insights
В каталоге вашей программы

Инвестируйте в себя

Новые знания — в удобное для вас время.

Начать на Coursera

Обучение откроется на Coursera
в новой вкладке

Обучение на Coursera

≈ 13.9 ч

9 модулей

Язык: Английский

Субтитры: Арабский, Французский, Итальянский, Бразильский португальский, Корейский, Немецкий, Пушту, Индонезийский, Испанский, Дари, Японский

Часть программы вашего университета
Ingestion + Cleaning: From S3 Logs to Partitioned ML DataВидео
Hands-on Activity: Build and Debug an Airflow + Spark ETL PipelineЗадание
02Engineer, Validate, and Govern ML Data: Ensuring Data Quality, Lineage, and Governance Across ML Pipelines8 материалов
Why Data Quality and Governance Matter for MLВидеоWhere Does Data Drift Show Up in Your Work?DIALOGUEWhat to Check: Dimensions of Data Quality and LineageЧтениеDetecting Drift and Preparing for AuditВидеоHands-on Activity: Validate Quality and Update Lineage After Schema DriftЗаданиеChoosing Effective Lineage Documentation PatternsЧтениеEnd-to-End Pipeline Validation LabЛабораторнаяGraded Quiz: Final Mastery CheckЗадание
03Transform and Communicate AI Insights Visually: Transforming Data for Insight 9 материалов
Welcome and Introduction ВидеоWhen Has Messy Data Slowed Down Your Work?DIALOGUEData Cleaning and Data TransformationЧтениеJoining CRM and Usage Tables: What You Need to Know FirstВидеоPandas Walkthrough: From Raw Tables to 30-Day AggregatesВидеоSQL vs. Pandas: Why Use SQL Over Pandas and Vice VersaЧтениеHands-On Activity: Transform a Mini-Dataset Using SQL or PandasЗаданиеQuiz: Data Joins, Aggregations, and Transformation ConceptsЗаданиеBuild a 30-Day Aggregated Dataset and Export ParquetЛабораторная
04Transform and Communicate AI Insights Visually: Evaluate Findings and Communicating Insights8 материалов
What makes an insight persuasive in your workplace?DIALOGUEWhy Insight Communication Influences Decisions More Than Data AloneВидеоEvaluating Findings Against Hypotheses: A Simple FrameworkВидеоHow to Use Different Funnel Visualizations to Effectively Tell Your Data Analytics StoryЧтениеBuild a Clear Funnel View and Identify Drop-Off CausesВидеоUnveiling McKinsey's Communication Secrets: the Pyramid PrincipleЧтениеHands-On Activity: Build a Funnel Visualization and Write a Drop-Off Insight ЗаданиеGraded Quiz: Visualizing and Communicating AI-Driven InsightsЗадание
05Analyze, Engineer, and Boost AI ROI: Why EDA Shapes Strong Feature Engineering8 материалов
When Has EDA Surprised You?DIALOGUEWelcome & IntroductionВидеоWhy Feature Engineering Starts with the Right QuestionsВидеоHow to Use EDA to Improve Model Performance with Feature EngineeringЧтениеInterpreting EDA Signals — Segments, Trends, OutliersВидеоHands-on Activity: Identify Feature Opportunities from Segment EDAЗаданиеFeature Selection using Chi-Square TestЧтениеPractice Quiz: Interpreting EDA to Guide Feature Engineering Задание
06Analyze, Engineer, and Boost AI ROI: Connecting Model Performance to Business Impact 9 материалов
Why A/B Testing Connects Models to ROIВидеоEvaluating Model Performance — Lift, Confidence, and Checkout EffectsВидеоA/B Testing: Statistical Significance ExplainedЧтениеWhat Makes an A/B Result Actionable?DIALOGUEHands-on Activity: Interpret an A/B Test for a Ranking Model ЗаданиеCommon Development Pitfalls in A/B Testing and How to Avoid ThemЧтениеWhen Does an A/B Result Truly Drive a Decision?DIALOGUEBuild an EDA-Driven Feature Candidate List and Test Model ImpactЛабораторнаяGraded Quiz: Evaluate, Experiment, and Prove AI ImpactЗадание
07Deconstruct AI: Complex ML Problems: Break Down Complex ML Systems with Modular Thinking 8 материалов
Welcome: Why Decomposition Matters in MLВидеоWhen Has an ML System Felt Too Complex?DIALOGUEModular Thinking in ML: Core Concepts and BenefitsВидеоThe Essential Modules in ML SystemsЧтениеReal-Time Fraud Detection: System BreakdownВидеоDecompose a Real-Time Fraud Detection PipelineЛабораторнаяUnderstanding Data Flow and Latency in ML PipelinesВидеоHands-on Activity: Improve a Flawed ML Pipeline DiagramЗадание
08Deconstruct AI: Complex ML Problems: Turn System Ideas Into Clear ML Abstractions 6 материалов
What Makes an Effective ML Abstraction?Видео Which Abstraction Styles Fit Your Thinking?DIALOGUEHow Flowcharts, System Maps, and Pseudocode Work TogetherЧтениеFeature Store Read/Write Pattern: Architecture and PseudocodeВидеоHands-on Activity: Create a Minimal Abstraction for a Modular ML PipelineЗаданиеGraded Quiz: Design a Modular ML System + Abstraction PackageЗадание
09Project: Building and Evaluating an End-to-End ML Data Pipeline3 материалов
Why Reliable Data Pipelines Matter in AI SystemsЧтениеProject Requirements for a Machine Learning Data PipelineЧтениеBuild a Machine Learning Data Pipeline for Churn PredictionЗадание