Count data
Count data consist of nonnegative integer-valued observations representing numbers of events, items, or occurrences.
Ensemble Methods for Machine Learning
A count label is a target variable representing a discrete nonnegative integer count of events or items.
- How it is used
- Models event or item totals when regression targets are counts rather than continuous values.
- How deep it goes
- Contrasts continuous and count assumptions, then connects Poisson distributions, likelihood-based losses, and boosting objectives.
- Examples
- Uses doctor visits, insurance claims, urban counts, and casual Bike Sharing users with zero-heavy, right-skewed labels.
- What you can do after
- Apply it via PoissonRegressor or distribution-specific boosting losses for count-valued prediction tasks.
Read how Ensemble Methods for Machine Learning explains count data →
Machine Learning System Design
Data consisting of nonnegative integer-valued observations, often recorded over time as counts of events or units.
- How it is used
- Frames retail demand forecasting around discrete outcomes and inventory decisions.
- How deep it goes
- Defines integer-valued time series and motivates predictive distributions, but stops short of deriving specific count models.
- Examples
- Uses item-store-period demand, sold units, delivered units, stockouts, and service-level agreements as examples.
- What you can do after
- Evaluate count forecasts using metrics aligned with discrete demand distributions and inventory-service goals.
Read how Machine Learning System Design explains count data →
Related concepts
- mean absolute percentage errorcontrasts with · Mean absolute percentage error is the average absolute prediction error divided by the corresponding actual value, expressed as a percentage.
- Tweedie distributionaddressed by · The Tweedie distribution is a family of exponential-dispersion probability distributions that includes compound Poisson-gamma and related discrete or continuous cases.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.