> ML_ALGORITHM // BAYESIAN-STRUCTURAL-TIME-SERIES-BSTS_v1.0
Bayesian Structural Time Series (BSTS)
State space modeling framework popularized by Google for counterfactual causal inference and transparent time series trend-seasonality decomposition.
State Space Dynamic Estimationbayesian-probabilistichigh-intrinsicsmall (<1k)
Back to All AlgorithmsComputational Complexity
Training Complexity:O(mcmc_draws * (kalman_filter + spike_slab))
Inference Complexity:O(draws * timesteps)
Hardware Profile
CPU Friendly:Yes
Requires GPU:No
Memory Footprint:low
Interpretability & Data
Interpretability Tier:high-intrinsic
Training Data Needs:small (<1k)
Interpretability Assessment
Spike-and-slab prior outputs explicit posterior inclusion probabilities for synthetic control predictors.
Suitable Tasks & Supported Modalities
Suitable Tasks:
causal impact analysistime series forecastingtrend decomposition
Supported Modalities:
time-series
Implementing Libraries
Foundational Literature
Common Pitfalls & Warnings
- Control pool series must not be affected by the intervention itself, otherwise counterfactual estimation is invalidated