Moving averages is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Research Wiki · Methods Lab
Time-series analysis and forecasting
Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals.
Methods family
Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals.
Use methods only after defining the estimand, data structure, assumptions, validation plan and decision consequence.
Weighted moving averages is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Exponential smoothing is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Holt’s linear trend is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Holt-Winters method is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Seasonal decomposition is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Classical decomposition is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →STL decomposition is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Autoregressive models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Moving-average models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Spearman correlation measures monotonic association using ranks rather than raw values. It is useful when relationships are non-linear but consistently ordered or when data are ordinal.
ARIMA models a differenced time series using autoregressive and moving-average terms. Diagnostics should test residual structure, stability and whether seasonality or external drivers require extensions.
Seasonal ARIMA extends ARIMA with seasonal differencing and seasonal autoregressive and moving-average terms. It is suitable when repeated seasonal patterns remain after transformation.
ARIMAX is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Vector autoregression is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Vector error-correction models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Cointegration is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Granger causality is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
State-space models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Kalman filtering is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Dynamic regression is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Transfer-function models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Intervention analysis is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Interrupted time-series analysis is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Prophet forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Bayesian structural time series is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Gaussian-process forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Neural-network forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
LSTM forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Temporal convolutional networks is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Transformer-based forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Hierarchical forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Demand forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Bass diffusion model is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Gompertz growth model is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Logistic growth model is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
New-product adoption forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Scenario forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Forecast accuracy is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
MAPE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
WAPE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
RMSE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
MAE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Forecast bias is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Prediction intervals is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Backtesting is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Rolling-origin validation is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Browse linked method entries
Moving averages
Moving averages is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodWeighted moving averages
Weighted moving averages is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodExponential smoothing
Exponential smoothing is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodHolt’s linear trend
Holt’s linear trend is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodHolt-Winters method
Holt-Winters method is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodSeasonal decomposition
Seasonal decomposition is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodClassical decomposition
Classical decomposition is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodSTL decomposition
STL decomposition is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodAutoregressive models
Autoregressive models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodMoving-average models
Moving-average models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodARMA
Spearman correlation measures monotonic association using ranks rather than raw values. It is useful when relationships are non-linear but consistently ordered or when data are ordinal.
Open methodARIMA
ARIMA models a differenced time series using autoregressive and moving-average terms. Diagnostics should test residual structure, stability and whether seasonality or external drivers require extensions.
Open methodSeasonal ARIMA
Seasonal ARIMA extends ARIMA with seasonal differencing and seasonal autoregressive and moving-average terms. It is suitable when repeated seasonal patterns remain after transformation.
Open methodARIMAX
ARIMAX is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodVector autoregression
Vector autoregression is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodVector error-correction models
Vector error-correction models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodCointegration
Cointegration is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodGranger causality
Granger causality is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodState-space models
State-space models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodKalman filtering
Kalman filtering is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodDynamic regression
Dynamic regression is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodTransfer-function models
Transfer-function models is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodIntervention analysis
Intervention analysis is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodInterrupted time-series analysis
Interrupted time-series analysis is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodProphet forecasting
Prophet forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodBayesian structural time series
Bayesian structural time series is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodGaussian-process forecasting
Gaussian-process forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodNeural-network forecasting
Neural-network forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodLSTM forecasting
LSTM forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodTemporal convolutional networks
Temporal convolutional networks is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodTransformer-based forecasting
Transformer-based forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodHierarchical forecasting
Hierarchical forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodDemand forecasting
Demand forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodBass diffusion model
Bass diffusion model is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodGompertz growth model
Gompertz growth model is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodLogistic growth model
Logistic growth model is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodNew-product adoption forecasting
New-product adoption forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodScenario forecasting
Scenario forecasting is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodForecast accuracy
Forecast accuracy is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodMAPE
MAPE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodWAPE
WAPE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodRMSE
RMSE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodMAE
MAE is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodForecast bias
Forecast bias is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodPrediction intervals
Prediction intervals is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodBacktesting
Backtesting is a method within time-series analysis and forecasting. Methods that model trend, seasonality, autocorrelation and external drivers to produce forecasts and uncertainty intervals. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodRolling-origin validation
Rolling-origin validation is a statistical or analytical concept within time-series analysis and forecasting. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodNo entries match this search.