Skip to content
DeepForgeby svx
Advancedcausal-inferenceupliftexperimentationtreatment-effects

Causal Inference & Uplift

25 problems·~11h·4 stages

Go beyond correlation to what actually changes outcomes. You will estimate propensity scores, weight and trim samples for average treatment effects, adjust with causal forests and quasi-experiments, build uplift models with trusted calibration, and test time-aware causality with Granger and cointegration tools. Take it after Statistics Mastery.

Your progress

0/25 · 0%

Path readiness 0%Easy 0% · Medium 0% · Hard 0%

Continue path · Logistic Propensity Score

What you will learn

  • Estimate propensity scores and inverse-probability-weighted effects
  • Adjust for confounding with trimming, causal forests, and quasi-experiments
  • Build uplift models and read their cumulative gain and calibration
  • Test time-aware causality with Granger F-tests and cointegration

Before you start

1Propensity Scores and Weighting

0/7

Make treated and control groups comparable: logistic propensity scores, match-pair counts, overlap fractions, weight trimming, ATT via propensity odds, IPW average treatment effects, and ATT propensity weights.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepLogistic Regression, From ScratchEasy

2Adjustment and Quasi-Experiments

0/6

When randomisation is not available: the propensity Newton step, trimming fractions, causal-forest honesty splits, quantile treatment effects, fuzzy regression-discontinuity jumps, and double-machine-learning residual estimates.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepLogistic Regression, From ScratchEasy

3Uplift Modeling

0/6

Predict who benefits: relative uplift, uplift per segment, response-rate uplift by decile, two-model uplift, cumulative gain, and uplift calibration error.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepLogistic Regression, From ScratchEasy

4Time-Aware Causality

0/6

Causality with a clock: Granger-causality F statistics from series, cumulative impulse responses, cointegration spreads, Engle-Granger residual changes, VAR coefficient application, and uplift gain at a targeting fraction.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepLogistic Regression, From ScratchEasy

Capstone

Contest

Close the path with one real artifact from the catalog.

Statistics Speedrun

A timed pass over the descriptive core that the causal estimators build on.