Glossary
Term
Definition
API
Application Programming Interface - a way for programs to communicate
Basic Mathematical Settings
A suite of mathematiocal settings controlling how the forecasting model interprets data, selects inputs, and structures the prediction task. Set to automatic by default but can be manually configured.
Comma Separated Values
A plain-text file format used to store tabular data where data records are separated by lines, and the columns within each record are separated by commas.
Data Horizons
The point in time when historical or predicted values can be used in a forecast, accounting for delays in measurement, reporting, or data collection
Deterministic
Outcomes that are completely predictable and without randonmess given the same initial conditions and inputs
Events
Unpredictable or irregular intervals of time
Feature
A transformed or engineered version of an input variable that provides additional predictive information to the model.
Forecast Horizon
The length of time into the future for which predictions are made.
Maximum
The largest value in the dataset
Mean
The average value,the sum of all values divided by the total number of values
Mean Absolute Error
The average absolute difference between predicted and actual values.
Mean Absolute Percentage Error
The average absolute error expressed as a percentage of the actual values.
Mean Absolute Scaled Error
Normalizes the MAE by the in-sample MAE of a naïve forecast, useful for comparing across datasets.
Mean Bias Error
Indicates whether the model systematically over- or under-predicts.
Median Absolute Error
The median of absolute errors, less sensitive to anomalies than MAE.
Minimum
The smallest value in the dataset
Model Zoo
A dynamic repository containing all of the built models for an experiment
Multivariate Time-Series Analysis
The process of analysing multiple time-dependent variables together to identify patterns and relationships
Naïve Forecast
A simple forecasting method where the forecast for the next period is the actual value from the most recent period
Observation
A single recorded value of a variable.
Root Mean Squared Error
The square root of the average squared differences between predictions and actual values.
Root Mean Squared Scaled Error
Similar to RMSE but scaled relative to a naïve forecast, allowing comparison across series.
Sampling Period
Similar to Sampling Rate, the interval between concurrent timestamps
Seasonality
A predictable, reoccuring pattern of fluctuations that occur at fixed intervals within a year
Standard Deviation
A measure of how spread out values are from the mean. Where a lower SD indicates greater clustering of values around the mean
Stochastic
Outcomes that are purely based on random probability
Symmetric Mean Absolute Percentage Error
Accounts for both over- and under-predictions symmetrically.
Tangent Credential
A Bearer token used to authenicate valid Tangent users, also referred to as an API Key or Token
Variable
A feature that is liable to vary or change. Typically inconsistent or having a fixed pattern