Bald verfügbar
Data Science & Analytics
Finde die Antwort in unordentlichen Daten – und beweise sie.
Der komplette Kurs
- Module
- 71
- Themen
- 818
- Min. pro Lektion
- 15
#1
Arbeitgeber sehen KI und Big Data als am schnellsten wachsende Kompetenz bis 2030.
Was du danach kannst.
Unordentliche Daten mit SQL und Python bändigen
Etappe 1 · Foundations & Data Wrangling
Hypothesen testen, Unsicherheit quantifizieren
Etappe 2 · Statistics & Inference
Prognosen bauen, die du verteidigen kannst
Etappe 3 · Modeling & Forecasting
A/B-Tests designen und richtig lesen
Etappe 4 · Causality & Experiments
Funnels, Kohorten und Abwanderung diagnostizieren
Etappe 5 · Product & Business Analytics
Ergebnisse präsentieren, die Entscheidungen ändern
Etappe 6 · Decisions & Delivery
Sieben Etappen.Ein stetiger Aufstieg.
Von der unordentlichen Tabelle bis zur Empfehlung, die jemand absegnet.
71Module
in 7 Etappen
818Themen
≈ 12 pro Modul
≈205Stunden
in Lektionen à 15 Minuten
≈2Jahre
≈13Monate
≈7Monate
bis zum Ende des Kurses
- 1
Foundations & Data Wrangling
Unordentliche Daten, vertrauenswürdig gemacht
15 Module · 175 Themen · ≈ 44 Std.
- 2
Statistics & Inference
Was die Zahlen sagen können — und was nicht
13 Module · 151 Themen · ≈ 38 Std.
- 3
Modeling & Forecasting
Modelle, die erklären und vorhersagen
10 Module · 118 Themen · ≈ 30 Std.
- 4
Causality & Experiments
Ursache, nicht nur Korrelation
7 Module · 82 Themen · ≈ 21 Std.
- 5
Product & Business Analytics
Die Zahlen, auf denen ein Business läuft
11 Module · 128 Themen · ≈ 32 Std.
- 6
Decisions & Delivery
Von der Analyse zur Entscheidung
10 Module · 112 Themen · ≈ 28 Std.
- 7
Practicum & Capstone
Eine Entscheidung, mit Daten begründet
5 Module · 52 Themen · ≈ 13 Std.
Monate
0612180369024
MonateZiel · ≈ 2 Jahre · ≈ 13 Monate · ≈ 7 Monate
Stunden und Monate sind Schätzungen: eine 15-Minuten-Lektion pro Thema, jeden Tag.
Jedes Modul.Jedes Thema.
Modul- und Themen-Titel bleiben Englisch — die Sprache, in der das Fach arbeitet.
Etappe 1
Foundations & Data Wrangling
Unordentliche Daten, vertrauenswürdig gemacht
15 Module · 175 Themen
1Foundations of Data Science and Analytics10 Themen
- Data Science, Analytics, and Business Intelligence
- Descriptive, Diagnostic, Predictive, and Prescriptive Questions
- Statistical Inference and Causal Inference
- Data as Evidence About a Larger Process
- From Raw Observations to Decisions
- The Analytical Project Lifecycle
- Domain Knowledge and Analytical Judgment
- Roles of Analysts, Data Scientists, and Data Engineers
- Reproducibility, Transparency, and Responsible Interpretation
- Recognizing What Available Data Cannot Answer
2Analytical Problem Framing11 Themen
- Translating a Business Question into an Analytical Question
- Defining the Decision an Analysis Will Inform
- Units of Observation and Units of Analysis
- Target Populations and Relevant Time Horizons
- Outcomes, Exposures, Treatments, and Covariates
- Estimands and Precisely Defined Quantities of Interest
- Hypotheses, Assumptions, and Competing Explanations
- Baselines, Comparators, and Success Criteria
- Data Availability and Feasibility Assessment
- Stakeholder Alignment and Scope Boundaries
- Choosing an Analytical Deliverable for the Decision
3Measurement and Data-Generating Processes12 Themen
- How Data Are Created, Recorded, and Selected
- Constructs, Proxies, and Operational Definitions
- Nominal, Ordinal, Interval, and Ratio Scales
- Continuous, Discrete, Binary, and Count Variables
- Measurement Validity and Reliability
- Measurement Error and Instrument Limitations
- Observation Windows and Exposure Time
- Units, Denominators, and Population at Risk
- Selection Mechanisms and Unobserved Cases
- Administrative Data and Designed Research Data
- Changes in Measurement Across Time and Groups
- Distinguishing Real-World Change from Recording Change
4Quantitative Foundations for Data Analysis12 Themen
- Ratios, Rates, Proportions, and Percentages
- Percentage Change and Percentage-Point Change
- Weighted Averages and Aggregation Rules
- Logarithms, Exponentials, and Growth Rates
- Functions and Transformations of Variables
- Vectors, Matrices, and Tabular Representations
- Dot Products and Linear Combinations
- Derivatives as Rates of Change
- Optimization Objectives and Loss Functions
- Numerical Precision and Rounding Effects
- Dimensional Consistency and Unit Conversion
- Order-of-Magnitude Checks and Plausibility Tests
5Spreadsheet Analytics and Analytical Prototyping12 Themen
- Structured Tables and Consistent Column Types
- Relative, Absolute, and Structured References
- Conditional Aggregation and Lookup Functions
- Pivot Tables and Grouped Summaries
- Array Calculations and Reusable Formulas
- Dates, Number Formats, and Hidden Type Errors
- Formula Auditing and Error Propagation
- Data Validation and Protected Input Areas
- Scenario Tables and What-If Models
- Reconciling Spreadsheet Totals with Source Records
- Documenting Manual Steps and Workbook Assumptions
- Choosing Between Spreadsheet, SQL, and Code-Based Workflows
6Analytical Programming with Python and R13 Themen
- Notebooks, Scripts, and Interactive Analysis
- Variables, Data Types, and Collections
- Arrays, Series, Data Frames, and Tables
- NumPy, pandas, and R Data Analysis Workflows
- Indexing, Filtering, and Boolean Logic
- Vectorized Operations and Elementwise Calculations
- Functions for Reusable Analytical Steps
- Iteration and Simulation Workflows
- Dates, Strings, and Categorical Data
- Missing Values and Type Conversion
- Reading Errors and Debugging Analytical Code
- Assertions and Checks on Intermediate Results
- Choosing and Documenting an Analysis Environment
7Tabular Data Transformation12 Themen
- Tidy Data and Consistent Table Structure
- Selecting, Renaming, and Reordering Columns
- Filtering Records and Deriving Variables
- Grouping and Aggregating Observations
- Reshaping Between Wide and Long Formats
- Joining Tables and Matching Keys
- Join Cardinality and Row Multiplication
- Concatenation, Unions, and Schema Alignment
- Sorting, Ranking, and Within-Group Operations
- Text Parsing and Structured Field Extraction
- Time-Based and Nearest-Match Joins
- Validating Transformations with Reconciliation Checks
8SQL Foundations for Analytics11 Themen
- Tables, Rows, Columns, and Relational Keys
- SELECT, WHERE, and ORDER BY
- Expressions, CASE Statements, and Derived Fields
- NULL Values and Three-Valued Logic
- COUNT, SUM, AVG, and Other Aggregations
- GROUP BY and HAVING
- Inner, Left, Full, and Cross Joins
- Subqueries and Common Table Expressions
- DISTINCT, Duplicates, and Counting Units
- Date, Time, and Interval Calculations
- Translating an Analytical Question into a Query
9Advanced Analytical SQL14 Themen
- Window Functions and Partitioned Calculations
- ROW_NUMBER, RANK, and Dense Ranking
- LAG, LEAD, and Event Sequences
- Running Totals and Rolling Windows
- Window Frames and Boundary Semantics
- Conditional Aggregation and Pivoted Summaries
- Querying Nested and Semistructured Data
- Recursive Queries for Hierarchical Relationships
- Sessionization and Time-Gap Logic
- Cohort Tables and Retention Queries
- Funnel Steps and Ordered Event Matching
- As-Of Logic and Historical State Reconstruction
- Query Plans and Expensive Join Patterns
- Verifying Complex Queries Against Small Known Examples
10Data Acquisition and Integration12 Themen
- Files, Databases, APIs, and Event Streams as Data Sources
- CSV, Spreadsheet, JSON, and Columnar Data Formats
- Encodings, Delimiters, and Schema Detection
- API Pagination, Rate Limits, and Incremental Retrieval
- Public Data Portals and Source Documentation
- Permission, Licensing, and Appropriate Data Reuse
- Identifiers and Cross-Source Record Matching
- Entity Resolution and Ambiguous Matches
- Timestamp Alignment and Timezone Conversion
- Source Reconciliation and Conflicting Values
- Preserving Raw Inputs and Acquisition Metadata
- Assessing Coverage Before Combining Datasets
11Analytical Data Modeling11 Themen
- Transactional and Analytical Data Structures
- Table Grain and the Meaning of One Row
- Fact Tables and Dimension Tables
- Star Schemas and Analytical Relationships
- Event Facts, Snapshots, and Accumulating Records
- Slowly Changing Dimensions and Historical Attributes
- Additive, Semiadditive, and Nonadditive Measures
- Many-to-Many Relationships and Bridge Tables
- Semantic Layers and Shared Metric Definitions
- Derived Tables and Reusable Analytical Datasets
- Preventing Double Counting Across Data Models
12Data Cleaning and Standardization11 Themen
- Profiling Types, Ranges, and Value Frequencies
- Parsing Numbers, Dates, and Structured Strings
- Standardizing Units, Categories, and Labels
- Duplicate Records and Duplicate Entities
- Impossible Values and Logical Contradictions
- Outliers, Rare Events, and Data Entry Errors
- Text Normalization and Identifier Cleanup
- Resolving Conflicting Records
- Handling Truncated, Rounded, and Censored Measurements
- Documenting Cleaning Decisions and Their Consequences
- Comparing Results Before and After Cleaning
13Missing Data and Measurement Error11 Themen
- Missing Completely at Random, at Random, and Not at Random
- Missingness Patterns and Plausible Mechanisms
- Complete-Case Analysis and Its Assumptions
- Simple Imputation and Distorted Uncertainty
- Multiple Imputation Principles
- Missingness Indicators and Their Interpretation
- Attrition, Dropout, and Lost Follow-Up
- Measurement Error in Outcomes and Predictors
- Misclassification of Categorical Variables
- Sensitivity Analysis for Unobserved Values
- Distinguishing Absence of an Event from Absence of a Record
14Data Quality and Lineage11 Themen
- Accuracy, Completeness, Consistency, and Timeliness
- Uniqueness, Validity, and Referential Integrity
- Data Contracts and Expected Schemas
- Field-Level and Dataset-Level Quality Checks
- Freshness, Volume, and Distribution Monitoring
- Reconciliation with Source Systems
- Lineage from Raw Data to Published Metrics
- Quality Ownership and Issue Escalation
- Schema Changes and Broken Assumptions
- Quantifying the Analytical Impact of Quality Problems
- Recording Known Limitations in Dataset Documentation
15Reproducible Analytical Workflows12 Themen
- Organizing Data, Code, Outputs, and Documentation
- Separating Raw, Intermediate, and Final Datasets
- Version Control for Analytical Work
- Environment and Dependency Reproducibility
- Random Seeds and Sources of Nondeterminism
- Parameterized Analyses and Repeatable Reports
- Executable Notebooks and Hidden State
- Lightweight Tests for Analytical Assumptions
- Peer Review of Queries, Code, and Conclusions
- Verifying AI-Assisted Code and Statistical Suggestions
- Recording Decisions, Exclusions, and Deviations
- Reproducing an Analysis from Its Documented Inputs
Etappe 2
Statistics & Inference
Was die Zahlen sagen können — und was nicht
13 Module · 151 Themen
16Descriptive Statistics12 Themen
- Counts, Frequencies, and Relative Frequencies
- Means, Medians, Modes, and Trimmed Means
- Variance, Standard Deviation, and Interquartile Range
- Quantiles, Percentiles, and Distribution Tails
- Skewness, Heavy Tails, and Multimodality
- Weighted Descriptive Statistics
- Rates with Unequal Exposure or Population Size
- Covariance and Correlation
- Cross-Tabulations and Conditional Summaries
- Aggregation Bias and Simpson's Paradox
- Robust Summaries for Skewed and Contaminated Data
- Choosing Summaries That Match the Analytical Question
17Exploratory Data Analysis11 Themen
- Establishing Dataset Structure and Coverage
- Exploring Individual Variable Distributions
- Examining Relationships Between Variables
- Comparing Groups and Subpopulations
- Detecting Anomalies and Unexpected Patterns
- Investigating Temporal and Spatial Variation
- Exploring Missingness and Data Collection Artifacts
- Testing Initial Explanations Against Alternative Views
- Separating Exploration from Confirmation
- Recording Hypotheses Generated During Exploration
- Identifying the Next Data or Analysis Needed
18Data Visualization Principles12 Themen
- Matching Chart Types to Analytical Tasks
- Visual Encodings and Perceptual Accuracy
- Distributions, Comparisons, Relationships, and Trends
- Scales, Axes, Baselines, and Transformations
- Small Multiples and Consistent Comparisons
- Visualizing Uncertainty and Sample Size
- Overplotting, Binning, and Density Displays
- Color, Contrast, and Accessible Design
- Annotations and Contextual Reference Lines
- Interactive Filtering and Linked Views
- Recognizing Misleading and Overdecorated Charts
- Choosing Tables When Exact Values Matter
19Probability and Conditional Reasoning11 Themen
- Events, Sample Spaces, and Probability Rules
- Conditional Probability and Independence
- Joint and Marginal Probabilities
- Bayes' Theorem and Updating Beliefs
- Base Rates and Inverse Probability Errors
- Independence and Conditional Independence
- Counting Rules and Combinatorial Reasoning
- Expected Value and Long-Run Averages
- Law of Total Probability and Total Expectation
- Translating Verbal Uncertainty into Probability Models
- Simulating Probability Problems to Check Intuition
20Random Variables and Probability Models12 Themen
- Discrete and Continuous Random Variables
- Probability Mass, Density, and Cumulative Distribution Functions
- Expectation, Variance, and Covariance
- Bernoulli and Binomial Models
- Poisson and Count Models
- Normal and Lognormal Models
- Exponential and Waiting-Time Models
- Heavy-Tailed and Mixture Distributions
- Joint Distributions and Dependence
- Transformations of Random Variables
- Laws of Large Numbers and the Central Limit Theorem
- Matching Distributional Assumptions to Observed Data
21Sampling and Study Design12 Themen
- Populations, Sampling Frames, and Samples
- Probability and Nonprobability Sampling
- Simple Random and Systematic Sampling
- Stratified and Cluster Sampling
- Multistage Sampling Designs
- Convenience Samples and Coverage Bias
- Selection Bias and Survivorship Bias
- Cross-Sectional, Longitudinal, and Repeated Cross-Sectional Studies
- Independent Observations and Clustered Observations
- Sampling Error and Nonsampling Error
- Generalizability and Target Population Alignment
- Designing Data Collection Around the Intended Inference
22Survey Analytics and Weighting11 Themen
- Questionnaire Design and Measurement Consistency
- Question Wording, Ordering, and Response Options
- Response Bias and Social Desirability
- Unit Nonresponse and Item Nonresponse
- Sampling Weights and Unequal Selection Probabilities
- Nonresponse Adjustments and Calibration
- Poststratification and Raking
- Design Effects and Effective Sample Size
- Weighted Estimation and Uncertainty
- Comparing Survey Responses with Behavioral Records
- Reporting Representation and Remaining Bias
23Estimation and Uncertainty11 Themen
- Parameters, Estimators, and Estimates
- Bias, Variance, and Mean Squared Error
- Sampling Distributions and Standard Errors
- Point Estimates and Interval Estimates
- Confidence Intervals and Their Interpretation
- Estimation for Means, Proportions, Rates, and Differences
- Maximum Likelihood as an Estimation Framework
- Uncertainty in Ratios and Nonlinear Quantities
- Prediction Intervals and Confidence Intervals
- Finite-Sample and Asymptotic Reasoning
- Communicating Uncertainty Without Hiding the Main Result
24Resampling and Computational Inference11 Themen
- Simulation as a Tool for Statistical Reasoning
- Bootstrap Resampling and Sampling Variability
- Bootstrap Confidence Intervals
- Permutation Tests and Exchangeability
- Randomization Tests and Assignment Mechanisms
- Resampling Paired and Clustered Observations
- Block Resampling for Dependent Temporal Data
- Monte Carlo Error and Simulation Precision
- Parametric and Nonparametric Resampling
- Situations Where Naive Resampling Fails
- Checking Analytical Results with Computational Experiments
25Hypothesis Testing14 Themen
- Null and Alternative Hypotheses
- Test Statistics and Reference Distributions
- P-Values and Their Interpretation Limits
- Type I and Type II Errors
- One-Sided and Two-Sided Tests
- Tests for Means, Proportions, and Group Differences
- Paired and Independent-Sample Comparisons
- Analysis of Variance and Planned Contrasts
- Contingency Tables and Categorical Association Tests
- Small-Sample and Exact Tests
- Rank-Based and Distribution-Free Methods
- Statistical Significance and Practical Importance
- Equivalence and Noninferiority Questions
- Reporting Effect Sizes Alongside Test Results
26Statistical Power and Sample Size Planning10 Themen
- Power as a Property of a Specified Study Design
- Minimum Detectable Effects and Decision-Relevant Effects
- Baseline Rates and Outcome Variability
- Sample Size for Means, Proportions, and Differences
- Allocation Ratios and Unequal Group Sizes
- Clustering and Repeated Measurements in Power Calculations
- Attrition and Incomplete Outcome Observation
- Duration Planning for Seasonal or Delayed Outcomes
- Simulation-Based Power Analysis
- Interpreting Inconclusive Results Without Post Hoc Power Claims
27Multiple Testing and Reliable Statistical Evidence12 Themen
- Families of Hypotheses and Multiplicity
- Family-Wise Error and False Discovery Rate
- Adjustment Methods and Their Assumptions
- Exploratory and Confirmatory Analysis Plans
- Selective Reporting and Researcher Degrees of Freedom
- P-Hacking and Repeated Unplanned Analyses
- Winner's Curse and Effect Size Inflation
- Replication and Independent Confirmation
- Sensitivity to Analytical Choices
- Multiverse and Specification-Curve Reasoning
- Preregistration and Transparent Deviations
- Evaluating Evidence Beyond a Significance Threshold
28Bayesian Analysis12 Themen
- Prior, Likelihood, and Posterior Distributions
- Bayesian Updating for Simple Data Models
- Weakly Informative and Domain-Informed Priors
- Prior Predictive Checks
- Posterior Summaries and Credible Intervals
- Posterior Predictive Distributions
- Bayesian Estimation of Rates and Group Differences
- Hierarchical Priors and Shrinkage
- Computational Posterior Approximation
- Convergence and Effective Sample Size Concepts
- Prior Sensitivity and Model Criticism
- Posterior Probabilities in Decision Contexts
Etappe 3
Modeling & Forecasting
Modelle, die erklären und vorhersagen
10 Module · 118 Themen
29Linear Regression and Explanatory Models12 Themen
- Simple and Multiple Linear Regression
- Least Squares and Fitted Relationships
- Coefficient Interpretation and Measurement Units
- Categorical Predictors and Reference Groups
- Interactions and Conditional Associations
- Transformations of Outcomes and Predictors
- Residual Variation and Explained Variation
- Standard Errors and Coefficient Intervals
- Conditional Mean and Exogeneity Assumptions
- Association, Adjustment, and Causal Interpretation
- Prediction for New Observations
- Communicating Regression Results in Practical Terms
30Generalized Linear Models12 Themen
- Outcome Distributions and Link Functions
- Logistic Regression for Binary Outcomes
- Odds, Odds Ratios, and Predicted Probabilities
- Marginal Effects and Standardized Predictions
- Multinomial and Ordinal Outcome Models
- Poisson Regression for Counts
- Offsets and Unequal Exposure
- Overdispersion and Negative Binomial Models
- Positive Skewed Outcomes and Gamma Models
- Excess Zeros and Two-Part Outcome Processes
- Likelihood-Based Estimation and Model Comparison
- Matching the Model to the Data-Generating Process
31Regression Diagnostics and Robust Analysis12 Themen
- Residual Patterns and Model Misspecification
- Heteroskedasticity and Robust Standard Errors
- Multicollinearity and Unstable Coefficients
- Leverage, Influence, and Unusual Observations
- Nonlinear Relationships and Flexible Functional Forms
- Splines and Generalized Additive Model Concepts
- Robust Regression and Sensitivity to Outliers
- Quantile Regression and Distributional Effects
- Regularization as a Tool for Stability
- Model Selection and Post-Selection Uncertainty
- Extrapolation and Unsupported Predictions
- Comparing Conclusions Across Reasonable Model Specifications
32Hierarchical, Longitudinal, and Panel Data11 Themen
- Repeated Observations and Within-Unit Dependence
- Nested and Crossed Data Structures
- Fixed Effects and Unit-Specific Baselines
- Random Intercepts and Random Slopes
- Partial Pooling Across Groups
- Within-Group and Between-Group Associations
- Cluster-Robust Inference
- Generalized Estimating Equation Concepts
- Time-Varying Covariates and Dynamic Outcomes
- Unequal Follow-Up and Unbalanced Panels
- Choosing Models for Conditional and Population-Average Questions
33Predictive Analytics and Model Validation13 Themen
- Prediction Targets and Decision Context
- Baseline Models and Benchmark Performance
- Training, Validation, and Test Data Roles
- Cross-Validation and Appropriate Data Splits
- Group-Based and Time-Based Validation
- Target Leakage and Information Available at Prediction Time
- Regression Error Metrics and Asymmetric Costs
- Classification Metrics and Class Imbalance
- Probability Calibration and Decision Thresholds
- Performance Variation Across Subpopulations
- Predictive Uncertainty and Distribution Shift
- Explaining Predictions Without Inferring Causation
- Comparing Model Utility with a Simple Decision Baseline
34Multivariate Analysis and Data Segmentation11 Themen
- Correlation Structure Across Multiple Variables
- Standardization and Scale-Sensitive Comparisons
- Principal Components as Low-Dimensional Summaries
- Explained Variance and Component Interpretation
- Factor Analysis and Latent Construct Assumptions
- Distances and Similarity Measures
- Clustering as an Exploratory Summary
- Choosing and Evaluating Segment Granularity
- Cluster Stability and Sensitivity to Preprocessing
- Profiling Segments with Independent Variables
- Distinguishing Discovered Groups from Natural Categories
35Time-Series Data and Temporal Structure11 Themen
- Time Indexes, Frequencies, and Observation Intervals
- Trends, Seasonality, Cycles, and Irregular Variation
- Calendar Effects and Timezone Boundaries
- Autocorrelation and Partial Autocorrelation
- Stationarity and Structural Change
- Differencing and Seasonal Adjustment
- Decomposition of Temporal Patterns
- Lagged Variables and Time Alignment
- Irregular Sampling and Temporal Aggregation
- Missing Periods and Revised Observations
- Distinguishing Leading Indicators from Lagged Correlations
36Statistical Forecasting12 Themen
- Forecast Horizons and Forecasting Objectives
- Naive, Seasonal Naive, and Drift Baselines
- Moving Averages and Exponential Smoothing
- Trend and Seasonal Exponential Smoothing Models
- Autoregressive and Moving-Average Components
- Differencing and ARIMA Models
- Seasonal ARIMA and Repeated Temporal Patterns
- Dynamic Regression with External Predictors
- Predictor Availability at the Forecast Origin
- Prediction Intervals and Forecast Distributions
- Combining Forecasts and Incorporating Structured Judgment
- Handling Structural Breaks and Regime Changes
37Forecast Evaluation and Demand Planning12 Themen
- Rolling-Origin and Expanding-Window Evaluation
- Horizon-Specific Forecast Accuracy
- MAE, RMSE, and Scaled Error Measures
- Percentage Error Metrics and Zero-Value Problems
- Forecast Bias and Systematic Underprediction
- Interval Coverage and Probabilistic Forecast Evaluation
- Hierarchical and Grouped Forecast Reconciliation
- Intermittent Demand and Sparse Outcomes
- Demand Censoring and Stockout Distortion
- Scenario Forecasts and Conditional Assumptions
- Forecasts, Targets, and Capacity Constraints
- Translating Forecast Error into Decision Consequences
38Survival and Event-History Analysis12 Themen
- Time-to-Event Outcomes and Observation Windows
- Right, Left, and Interval Censoring
- Truncation and Delayed Entry
- Survival, Hazard, and Cumulative Hazard Functions
- Kaplan-Meier Estimation
- Comparing Survival Curves
- Cox Proportional Hazards Models
- Proportional Hazards Assumptions and Diagnostics
- Accelerated Failure Time Models
- Competing Risks and Cumulative Incidence
- Restricted Mean Survival Time
- Event Recurrence and Time-Varying Exposures
Etappe 4
Causality & Experiments
Ursache, nicht nur Korrelation
7 Module · 82 Themen
39Foundations of Causal Inference12 Themen
- Counterfactual Questions and Potential Outcomes
- Average and Conditional Treatment Effects
- Consistency and Well-Defined Interventions
- Exchangeability and Confounding
- Positivity and Treatment Overlap
- Causal Diagrams and Directed Acyclic Graphs
- Confounders, Mediators, and Colliders
- Backdoor Paths and Adjustment Sets
- Selection Bias and Conditioning on Post-Treatment Variables
- Identification Versus Estimation
- Interference and Spillover Effects
- Generalization and Transport to New Populations
40Causal Inference from Observational Data12 Themen
- Emulating a Target Trial with Observational Records
- Defining Eligibility, Treatment, Follow-Up, and Outcomes
- Outcome Regression and Standardization
- Propensity Scores and Their Interpretation
- Matching and Covariate Balance
- Inverse Probability Weighting
- Extreme Weights and Limited Overlap
- Doubly Robust Estimation Concepts
- Time-Varying Treatments and Confounding
- Negative Controls and Falsification Checks
- Sensitivity to Unmeasured Confounding
- Reporting Assumptions That Data Alone Cannot Verify
41Natural Experiments and Identification Strategies11 Themen
- Natural Experiments and Plausibly Exogenous Variation
- Instrumental Variables and the Assignment Mechanism
- Relevance, Independence, and Exclusion Restrictions
- Local Average Treatment Effects and Monotonicity
- Weak Instruments and Unstable Estimates
- Regression Discontinuity Around Assignment Thresholds
- Sharp and Fuzzy Discontinuity Designs
- Bandwidth Selection and Local Comparisons
- Manipulation, Sorting, and Continuity Checks
- Placebo Thresholds and Alternative Explanations
- Limits of Generalizing Local Causal Effects
42Panel Data Methods for Causal Analysis11 Themen
- Before-After Comparisons and Their Limitations
- Difference-in-Differences Design
- Parallel Trends and the Untreated Counterfactual
- Event Studies and Dynamic Treatment Effects
- Anticipation and Treatment Timing
- Staggered Adoption and Heterogeneous Effects
- Serial Correlation and Appropriate Inference
- Synthetic Control Design and Donor Selection
- Pre-Treatment Fit and Placebo Comparisons
- Interrupted Time Series and Concurrent Changes
- Sensitivity to Comparison Groups and Time Windows
43Randomized Experimental Design12 Themen
- Random Assignment and Causal Identification
- Experimental Units and Analysis Units
- Control Groups and Treatment Conditions
- Blocking, Stratification, and Matched Designs
- Allocation Concealment and Blinding Concepts
- Primary Outcomes and Predefined Analysis Plans
- Noncompliance and Intention-to-Treat Effects
- Attrition and Missing Outcomes in Experiments
- Balance Checks and Interpretation Limits
- Experiment Duration and Delayed Effects
- Ethical Constraints and Participant Impacts
- Connecting Experimental Results to the Target Decision
44Online Controlled Experiments12 Themen
- A/B Testing in Digital Products
- Eligibility, Assignment, Exposure, and Analysis Populations
- Stable Assignment and Cross-Device Identity
- A/A Tests and Instrumentation Validation
- Sample Ratio Mismatch and Allocation Failures
- Primary Metrics, Guardrails, and Diagnostic Metrics
- Triggered Analyses and Exposure Bias
- Novelty, Learning, and Carryover Effects
- Bots, Internal Users, and Invalid Traffic
- Attribution Windows and Delayed Outcomes
- Experiment Interactions and Concurrent Changes
- Interpreting and Communicating a Product Experiment
45Advanced Experimentation and Treatment Heterogeneity12 Themen
- Factorial Designs and Interaction Effects
- Multivariant Experiments and Comparison Families
- Cluster Randomization and Group-Level Outcomes
- Switchback Designs and Time-Based Assignment
- Network Effects and Interference-Aware Experiments
- Sequential Testing and Planned Monitoring
- Alpha Spending and Repeated-Look Error Control
- Bayesian Monitoring and Decision Rules
- CUPED and Pre-Experiment Covariate Adjustment
- Heterogeneous Treatment Effects and Subgroup Credibility
- Adaptive Allocation and Inference Tradeoffs
- Long-Term Holdouts and Persistent Effects
Etappe 5
Product & Business Analytics
Die Zahlen, auf denen ein Business läuft
11 Module · 128 Themen
46Metric Design and Metric Systems12 Themen
- Business Objectives and Operational Metrics
- North Star Metrics and Supporting Measures
- Metric Trees and Driver Decomposition
- Leading, Lagging, and Guardrail Metrics
- Numerator, Denominator, and Eligibility Definitions
- Users, Accounts, Sessions, Events, and Transactions
- Ratios of Totals and Averages of Ratios
- Aggregation Windows and Timezone Policies
- Metric Sensitivity and Responsiveness
- Goodhart's Law and Metric Gaming
- Ownership, Versioning, and Metric Change Governance
- Distinguishing a Useful Proxy from the Desired Outcome
47Event Instrumentation and Behavioral Data12 Themen
- Tracking Plans and Event Taxonomies
- Event Names, Properties, and Schema Contracts
- Client-Side and Server-Side Measurement
- Event Time, Processing Time, and Late Arrival
- Identity Resolution and Anonymous-to-Known Transitions
- Deduplication, Retries, and Idempotent Event Handling
- Session Boundaries and Activity Windows
- Consent, Blocking, and Unobserved Behavior
- Attribution Metadata and Source Context
- Instrumentation QA and End-to-End Reconciliation
- Tracking Changes Across Application Releases
- Identifying Measurement Gaps Before Interpreting Behavior
48Product and Engagement Analytics11 Themen
- Acquisition, Activation, Engagement, and Value Delivery
- Defining Meaningful Active Use
- Feature Adoption and Depth of Use
- Frequency, Recency, and Engagement Intensity
- Active User Ratios and Their Limitations
- User Journeys and Behavioral Sequences
- Exposure Opportunity and Actual Feature Usage
- Account-Level and User-Level Product Analysis
- Segment Mix and Apparent Engagement Changes
- Diagnosing Product Friction with Behavioral Evidence
- Linking Product Metrics to User and Business Outcomes
49Funnel and Conversion Analysis11 Themen
- Funnel Steps and Eligible Starting Populations
- Open and Closed Funnel Definitions
- User-Level, Session-Level, and Event-Level Conversion
- Ordered and Unordered Step Completion
- Conversion Windows and Delayed Completion
- Repeated Attempts and Deduplicated Outcomes
- Step-Level Drop-Off and Time Between Steps
- Segment Comparisons with Consistent Denominators
- Cross-Device and Cross-Channel Journey Gaps
- Separating Traffic Mix Changes from Conversion Changes
- Testing Explanations for Funnel Bottlenecks
50Cohort, Retention, and Churn Analysis12 Themen
- Acquisition, Activation, and Behavioral Cohorts
- Calendar Time and Time Since Cohort Entry
- Fixed-Period, Rolling, and Return-Window Retention
- Cohort Maturity and Incomplete Observation
- Retention Curves and Retention Matrices
- Churn Definitions and Inactivity Thresholds
- Reactivation and Resurrected Users
- Customer Churn and Revenue Churn
- Survival-Based Interpretation of Retention
- Comparing Cohorts Under Changing Product and Channel Mix
- Leading Signals and Causal Claims About Churn
- Long-Term Retention with Limited Follow-Up
51Customer, Revenue, and Unit Economics Analytics13 Themen
- Customer, Account, Subscription, and Transaction Units
- Revenue Definitions and Recognition Timing
- Bookings, Billings, Cash Receipts, and Reported Revenue
- Recurring and Nonrecurring Revenue
- Gross and Net Revenue Retention
- Revenue per User and Revenue per Paying Customer
- Discounts, Refunds, and Failed Payments
- Contribution Margin and Cost Allocation Assumptions
- Customer Acquisition Cost and Attribution Scope
- Customer Lifetime Value and Observation Horizon
- Payback Period and Cohort Economics
- Discounting, Uncertainty, and Scenario-Based Value Estimates
- Avoiding Misleading Averages Across Customer Segments
52Marketing Measurement and Incrementality12 Themen
- Campaign Taxonomies and Acquisition Source Definitions
- Reach, Frequency, Response, and Conversion Metrics
- Attribution Windows and Touchpoint Visibility
- First-Touch, Last-Touch, and Multitouch Attribution
- Attribution Credit and Incremental Impact
- Selection Bias in Channel Comparisons
- Holdout Tests and Conversion Lift Studies
- Geographic Experiments and Market-Level Comparisons
- Marketing Mix Modeling and Identification Limits
- Carryover, Saturation, and Channel Interactions
- Organic and Paid Activity Overlap
- Evaluating Measurement Under Incomplete Tracking
53Operational and Process Analytics11 Themen
- Process Events, Cases, and Activity Sequences
- Throughput, Cycle Time, and Waiting Time
- Capacity, Utilization, and Bottlenecks
- Queueing Concepts and Little's Law
- Service Levels and Tail Performance
- Demand Variability and Workload Planning
- Process Mining and Deviations from Expected Flows
- Rework, Error Rates, and First-Pass Outcomes
- Resource Productivity and Case-Mix Adjustment
- Operational Changes and Causal Evaluation
- Turning Process Findings into Testable Improvement Proposals
54Geospatial Analytics11 Themen
- Coordinates, Geometries, and Spatial Data Types
- Coordinate Reference Systems and Projections
- Geocoding and Location Uncertainty
- Spatial Joins and Geographic Aggregation
- Distances, Areas, and Neighborhood Definitions
- Spatial Autocorrelation and Dependence
- Hotspots and Local Concentrations
- Rates, Population Denominators, and Exposure
- Modifiable Areal Units and Ecological Fallacies
- Spatial Sampling and Uneven Data Coverage
- Communicating Geographic Patterns Without Overclaiming Causes
55Text and Qualitative Data Analytics12 Themen
- Text as Unstructured Analytical Evidence
- Document Units and Corpus Construction
- Text Cleaning and Language Identification
- Word, Phrase, and Document Frequency Summaries
- Search Terms, Dictionaries, and Coding Rules
- Human Coding and Annotation Guidelines
- Inter-Rater Agreement and Label Quality
- Sentiment Measures and Contextual Limitations
- Topic Summaries and Independent Validation
- Linking Qualitative Themes to Quantitative Records
- Representative Quotations and Selective Evidence Risk
- Evaluating Automated Text Outputs Against Human Review
56Graph and Network Analytics11 Themen
- Nodes, Edges, and Relationship Definitions
- Directed, Undirected, and Weighted Networks
- Bipartite and Multilayer Data Structures
- Degree, Connectivity, and Component Structure
- Centrality Measures and Their Interpretation
- Communities and Group Structure
- Paths, Reachability, and Network Distance
- Temporal Networks and Relationship Change
- Sampling Bias and Missing Edges
- Distinguishing Network Association from Social Influence
- Visualizing Networks Without Obscuring Their Structure
Etappe 6
Decisions & Delivery
Von der Analyse zur Entscheidung
10 Module · 112 Themen
57Anomaly, Change, and Metric Monitoring11 Themen
- Baseline Behavior and Expected Variation
- Point Anomalies and Contextual Anomalies
- Control Charts and Stable Process Assumptions
- Seasonality-Aware Monitoring
- Change Points and Structural Shifts
- Threshold Selection and Alert Burden
- Base Rates and False Alarm Interpretation
- Multiple Metrics and Repeated Monitoring
- Decomposing an Unexpected Metric Movement
- Distinguishing Data Failures from Business Changes
- Confirming a Signal Before Recommending Action
58Simulation and Scenario Analysis11 Themen
- Defining a System and Its Decision Variables
- Deterministic and Stochastic Scenarios
- Monte Carlo Simulation of Uncertain Outcomes
- Choosing Distributions for Uncertain Inputs
- Correlated Inputs and Joint Scenarios
- Discrete-Event Simulation Concepts
- Queueing and Capacity Simulation
- Sensitivity Analysis and Dominant Assumptions
- Stress Scenarios and Tail Outcomes
- Calibration and Validation Against Observed Behavior
- Reporting Simulation Uncertainty and Model Limitations
59Optimization and Prescriptive Analytics11 Themen
- Decision Variables, Objectives, and Constraints
- Linear Programming Formulations
- Integer and Mixed-Integer Decision Models
- Resource Allocation and Scheduling Problems
- Assignment, Routing, and Network Flow Concepts
- Constraint Feasibility and Conflicting Requirements
- Sensitivity, Shadow Prices, and Marginal Value
- Multiple Objectives and Pareto Tradeoffs
- Stochastic and Robust Optimization Concepts
- Validating Optimized Decisions Under Real-World Conditions
- Recognizing When Model Assumptions Dominate the Recommendation
60Decision Analysis and Value of Information11 Themen
- Decisions, States of the World, and Consequences
- Expected Value and Expected Utility
- Decision Trees and Probabilistic Outcomes
- Asymmetric Costs and Decision Thresholds
- Risk Preferences and Downside Constraints
- Regret and Robust Decision Alternatives
- Value of Perfect and Partial Information
- Whether Additional Data Are Worth Collecting
- Combining Causal Effects with Costs and Constraints
- Sensitivity of the Preferred Decision to Assumptions
- Communicating a Recommendation Under Uncertainty
61Business Intelligence and Dashboard Design12 Themen
- Executive, Analytical, and Operational Dashboards
- Audience Needs and Decision Frequency
- Metric Hierarchies and Information Priorities
- Consistent Definitions Across Reports
- Filters, Drill-Downs, and Interaction Design
- Time Comparisons and Reference Periods
- Alerts, Exceptions, and Actionable Views
- Displaying Uncertainty and Data Freshness
- Avoiding Misleading Cross-Filter Calculations
- Dashboard Performance and Usability
- Testing Whether a Dashboard Supports Its Intended Decisions
- Retirement of Unused and Redundant Reports
62Analytical Communication and Data Storytelling12 Themen
- Leading with the Question and Main Finding
- Separating Evidence, Interpretation, and Recommendation
- Structuring an Analytical Narrative
- Choosing the Right Level of Technical Detail
- Explaining Effect Sizes in Familiar Units
- Presenting Uncertainty and Alternative Explanations
- Communicating Null and Inconclusive Results
- Avoiding Causal Language for Associational Findings
- Executive Summaries and Decision Memos
- Technical Appendices and Reproducible Evidence
- Responding to Challenges and Revising Conclusions
- Visual and Verbal Consistency Across Deliverables
63Efficient Analytics at Scale11 Themen
- Memory, Compute, Storage, and Data Movement Costs
- Columnar Data and Selective Reading
- Partitioning and Predicate Pushdown
- Query Optimization for Analytical Workloads
- Join Strategy, Data Skew, and Intermediate Result Size
- Chunked and Out-of-Core Processing
- Vectorization and Avoiding Unnecessary Recalculation
- Sampling and Approximate Aggregation
- Accuracy-Cost Tradeoffs in Approximate Results
- Distributed Processing Semantics Relevant to Analysts
- Validating That Faster Computation Preserves the Intended Result
64Data Governance and Responsible Access11 Themen
- Data Ownership, Stewardship, and Accountability
- Catalogs, Dictionaries, and Business Definitions
- Data Provenance, Licensing, and Reuse Conditions
- Access According to Analytical Need
- Sensitive Attributes and Data Minimization
- Pseudonymization and Reidentification Risk
- Privacy-Preserving Aggregation and Disclosure Risk
- Retention, Deletion, and Reproducibility Tensions
- Sharing Data, Code, and Results Responsibly
- Documenting Approved Uses and Known Limitations
- Coordinating Analysis with Privacy and Security Requirements
65Fairness, Ethics, and Societal Impact11 Themen
- Representation and Historical Bias in Data
- Measurement Bias and Unequal Error
- Proxy Variables and Sensitive Characteristics
- Subgroup Evaluation and Small-Sample Uncertainty
- Competing Fairness Definitions and Decision Context
- Feedback Loops Created by Analytical Decisions
- Who Benefits and Who Bears the Costs
- Consent, Expectations, and Appropriate Data Use
- Human Oversight and Contestability
- Ethical Review of High-Impact Analyses
- Communicating Limitations Without Overstating Objectivity
66Analytics Delivery, Monitoring, and Maintenance11 Themen
- Delivering Reusable Datasets, Reports, and Analytical Models
- Ownership and Handoff Documentation
- Scheduled Refreshes and Data Dependency Management
- Versioned Logic and Historical Result Reproducibility
- Acceptance Checks Before Publishing Results
- Monitoring Freshness, Completeness, and Metric Behavior
- Investigating Failed Refreshes and Inconsistent Outputs
- Upstream Changes and Analytical Drift
- Revising Models and Metrics with Controlled Comparisons
- Communicating Corrections to Published Findings
- Retiring Analyses That No Longer Support Valid Decisions
Etappe 7
Practicum & Capstone
Eine Entscheidung, mit Daten begründet
5 Module · 52 Themen
67Data Preparation and SQL Practicum10 Themen
- Turning an Ambiguous Request into a Dataset Specification
- Profiling and Cleaning a Messy Multisource Dataset
- Resolving Identifier and Join Cardinality Problems
- Building an Analytical Table at the Correct Grain
- Implementing Event Windows and Historical State Logic
- Creating Cohort, Funnel, and Retention Queries
- Reconciling Metrics Across Independent Calculations
- Investigating Missing Data and Measurement Gaps
- Documenting Data Lineage and Quality Limitations
- Delivering a Reproducible Preparation Workflow
68Statistical Inference and Modeling Practicum10 Themen
- Describing a Distribution Without Hiding Important Variation
- Designing a Sample and Assessing Representation
- Estimating an Effect with Appropriate Uncertainty
- Comparing Parametric and Resampling-Based Results
- Interpreting Multiple Tests Without Selective Reporting
- Fitting and Critiquing a Regression Model
- Analyzing a Binary or Count Outcome
- Evaluating Repeated or Clustered Observations
- Comparing Predictive Performance on Appropriate Held-Out Data
- Writing a Statistical Findings Memo with Explicit Assumptions
69Experimentation and Causal Analysis Practicum10 Themen
- Defining a Causal Question and Drawing Its Assumptions
- Planning an Experiment with Decision-Relevant Power
- Auditing Assignment, Exposure, and Outcome Measurement
- Diagnosing Sample Ratio Mismatch and Attrition
- Analyzing an A/B Test with Primary and Guardrail Metrics
- Evaluating the Credibility of a Subgroup Effect
- Assessing Covariate Balance in an Observational Study
- Critiquing a Difference-in-Differences or Discontinuity Design
- Conducting Sensitivity Analysis for a Causal Claim
- Preparing an Experiment Readout and Decision Recommendation
70Forecasting and Decision Analytics Practicum10 Themen
- Preparing a Time Series with Calendar and Missing-Period Checks
- Establishing Naive and Seasonal Forecast Baselines
- Comparing Forecasts Through Rolling-Origin Evaluation
- Assessing Prediction Intervals and Tail Risk
- Reconciling Forecasts Across Related Business Units
- Diagnosing Retention and Revenue Changes by Cohort
- Simulating an Operational Decision Under Uncertainty
- Formulating a Constrained Resource Allocation Problem
- Evaluating Whether More Information Would Change the Decision
- Presenting a Recommendation with Sensitivity to Key Assumptions
71Integrated Data Science and Analytics Capstone12 Themen
- Selecting a Decision and Defining an Answerable Question
- Specifying the Population, Unit, Time Horizon, and Estimand
- Auditing Data Sources and Measurement Processes
- Building and Validating an Analytical Dataset
- Exploring Patterns and Recording Competing Explanations
- Choosing Methods Appropriate to the Question and Data
- Quantifying Uncertainty and Testing Robustness
- Distinguishing Descriptive, Predictive, and Causal Conclusions
- Evaluating Practical Value, Fairness, and Decision Consequences
- Delivering a Clear Report or Decision-Support Dashboard
- Providing Reproducible Code, Documentation, and Review Evidence
- Defining Follow-Up Measurement and Maintenance Needs
Fünfzehn Minuten.Jeden Tag.
- 1
Eine Lektion passt in die Mittagspause
Eine Idee nach der anderen, auf kurzen Folien. Eine ganze Lektion dauert etwa fünfzehn Minuten.
- 2
Übung mit sofortigem Feedback
Die Fragen stecken in der Lektion. Antworte und sieh sofort, ob es sitzt.
- 3
Eine Serie, die dich zurückholt
Eine Lektion am Tag hält die Serie am Leben. Kleine, stetige Schritte tragen dich durch.
- 1
Eine Lektion passt in die Mittagspause
Eine Idee nach der anderen, auf kurzen Folien. Eine ganze Lektion dauert etwa fünfzehn Minuten.
- 2
Übung mit sofortigem Feedback
Die Fragen stecken in der Lektion. Antworte und sieh sofort, ob es sitzt.
- 3
Eine Serie, die dich zurückholt
Eine Lektion am Tag hält die Serie am Leben. Kleine, stetige Schritte tragen dich durch.
Für alle mit einer Kennzahl.
Analysten, die Spreadsheets entwachsen
Wo du am meisten wächst
Unordentliche Daten mit SQL und Python bändigen
Etappe 1 · Foundations & Data WranglingPMs, die eine Kennzahl verantworten
PhDs auf dem Sprung in die Industrie
Wo du am meisten wächst
Funnels, Kohorten und Abwanderung diagnostizieren
Etappe 5 · Product & Business Analytics
Wohin dieser Kurs führt.
Der Job, um den dieser Kurs gebaut ist, und wie Menschen hineinkommen.
Data Analyst
Macht aus den Daten eines Unternehmens Antworten, nach denen Menschen handeln können, und wächst in den Modellbau hinein.
Alle ZukunftsberufeIm Job
- Die Abfrage schreiben, die die Frage dieser Woche beantwortet
- Ein Dashboard bauen, das die Geschäftsleitung wirklich liest
- Ein Experiment entwerfen und sein Ergebnis richtig lesen
So kommt man hinein
Viele starten als Analyst:innen mit SQL, Tabellen und Statistik-Grundlagen; Data-Scientist-Rollen verlangen meist ein quantitatives Studium.
Sei von Anfang an dabei.
Frühzugriff für Einzelpersonen, Piloten für Teams. Sag uns, wer lernt.
enterprise@astratrainer.com