Technische GIDS

Voorspellingsnauwkeurigheidsstatistieken: MAPE, sMAPE en MASE

Forecast metrics summarize how predictions differ from observed values, but MAPE, sMAPE and MASE handle scale and zero values differently.

Luister naar deze gidsAI-verteld · ~3 min

Op deze pagina3 minuten lezen
  1. Overzicht
  2. Diepe duik
  3. Implementatie in de echte wereld
  4. The Future of Forecast Accuracy Metrics: MAPE, sMAPE and MASE
  5. Het in de praktijk brengen
  6. Blijf verkennen
  7. Veelgestelde vragen

Overzicht

Choose a metric that matches the decision, compare forecasts on the same evaluation windows and include a naive baseline rather than relying on a percentage score alone.

Diepe duik

Forecast accuracy metrics turn prediction errors into summaries, but each emphasizes different properties. Mean absolute error averages absolute errors in the target's units. MAPE averages absolute error divided by the absolute actual value, often multiplied by 100. It is easy to read as a percentage, but it is undefined when actual values are zero and can be dominated by small actuals. It also treats over- and under-forecast errors asymmetrically in some settings because actuals appear only in the denominator. sMAPE attempts a symmetric percentage by scaling absolute error relative to the magnitudes of actual and forecast. Several formulas exist, some use a factor of two and others not, and zero-over-zero cases require conventions. Despite its name, sMAPE can behave unexpectedly around zero and does not guarantee fair comparisons across all scales. Always state the exact implementation. Mean absolute scaled error (MASE) divides a model's test-set MAE by the in-sample MAE of a naive forecast, typically the one-step persistence forecast for nonseasonal data. For seasonal data, a seasonal-naive denominator may be more appropriate. A MASE below one means the model's average absolute error on the evaluation cases is lower than the chosen naive in-sample scale. It does not mean the model always beats the baseline on every observation, and the denominator can be zero for perfectly constant training data. Use an evaluation period that reflects the intended forecast horizon, preserve time order and compare models on identical cases. Report metric definitions, units, treatment of zeros, aggregation weights and baseline. If many series are combined, averaging percentage errors may give tiny-volume series disproportionate influence. Pair a headline metric with error distributions and business costs, such as stockouts versus overstock. No single metric captures calibration of prediction intervals, bias, tail risk and operational impact at once.

Implementatie in de echte wereld

A hypothetical forecast predicts 90 units when actual demand is 100. The absolute percentage error is 10%, using absolute error divided by the actual value.

When actual demand is zero, MAPE's denominator is zero and the percentage error is undefined; replacing zero with a small constant changes the metric and should be disclosed.

An analyst uses MASE and divides test-set MAE by the in-sample one-step naive MAE. A value below one means lower error than that naive scale on the evaluated cases, not guaranteed future superiority.

A team compares sMAPE implementation formulas before publishing results because definitions differ in denominator and scaling; values from different conventions may not be comparable.

The Future of Forecast Accuracy Metrics: MAPE, sMAPE and MASE

Forecast dashboards can be more honest by placing the metric formula, evaluation horizon, zero handling and naive baseline beside each result. Teams should show scale-dependent error in units as well as scale-free measures when comparing products or regions. As demand patterns shift, recompute the baseline using only information available at each forecast origin. Decision-weighted costs can complement statistical metrics when over- and under-forecasting have different consequences. Clear metric choices make it easier to compare models without implying that one percentage number captures every aspect of forecast quality.

Het in de praktijk brengen

Waarom het ertoe doet

  • Kosten en budget. Architectuurbeslissingen bepalen jarenlang de prestaties en bedrijfskosten.
  • Duidelijkere beslissingen. Technisch onderwijs helpt teams bij het kiezen van de juiste stapel, niet alleen de nieuwste.
  • Kwaliteitscontrole. Betere technische keuzes verminderen het aantal betrouwbaarheidsincidenten in de productie.

Pas op voor

  • Het optimaliseren van één benchmark kan bredere systeemzwakheden verbergen.
  • Infrastructuur- en onderhoudskosten worden vaak onderschat.
  • De lacunes op het gebied van beveiliging en waarneembaarheid kunnen groter worden naarmate systemen complexer worden.

Volgende stappen

  1. Definieer latentie-, kwaliteits- en kostendoelen vóór implementatie.
  2. Benchmark onder realistische belasting- en gegevensomstandigheden.
  3. Instrumentbewaking op fouten, drift en gebruikersimpact.
  4. Bereid rollback- en incidentresponspaden voor voordat u gaat schalen.

Blijf verkennen

Gratis nieuwsbrief

Ontvang de dagelijkse AI-briefing

Drie verifieerde AI-verhalen elke weekdagmorgen, geschreven in eenvoudig Engels. Gratis voor altijd, geen advertenties.

Eén e-mail elke weekdag. Afmelden met één klik. We verkopen of delen je adres nooit.

Test jezelf

Take the Forecast Accuracy Metrics: MAPE, sMAPE and MASE quiz

Directe feedback op elke antwoord, en een deelbare certificaat met een verifieerbare ID zodra je een cursus hebt afgerond.

Quiz starten

Ondersteun gratis AI-onderwijs. AI Understanding is een 501(c)(3) non-profit — geen advertenties, geen betaalmuur, nooit. Doe een donatie

Veelgestelde vragen

What is Forecast Accuracy Metrics: MAPE, sMAPE and MASE?

Forecast metrics summarize how predictions differ from observed values, but MAPE, sMAPE and MASE handle scale and zero values differently. Choose a metric that matches the decision, compare forecasts on the same evaluation windows and include a naive baseline rather than relying on a percentage score alone.

Actual demand is 100 and the forecast is 90. What is the absolute percentage error under MAPE?

Absolute error is 10, divided by actual 100, giving 0.10 or 10%.

What issue arises for MAPE when an actual value is zero?

MAPE divides by the actual magnitude, so a zero actual makes the percentage undefined.

What does MASE below one indicate under its usual definition?

MASE compares test MAE with a chosen naive in-sample scale; below one indicates lower average error than that scale.

Why state the exact sMAPE formula in a report?

Different sMAPE variants and zero-handling choices can produce noncomparable values.

Why can MAPE overemphasize low-volume observations?

Dividing by a small actual makes the same absolute error a larger percentage contribution.