Predictive Analytics icon

Predictive Analytics

Run explainable forecasting operations on finance time series without arbitrary formulas.

Actions4

Trend Forecast

AI-generated

Summary

Fit a linear trend over the sequential order of input time-series data rows and forecast a specified number of future points using simple linear regression.

Inputs

  • Currency — Optional currency metadata to include in the predictive result.
  • Forecast Horizon (required) — Number of future forecast points to generate, must be at least 1.
  • Max Recommended Horizon (required) — Recommended maximum forecast horizon. Larger values are allowed but produce warnings.
  • Value Column (required) — Name of the column containing the numeric financial series to forecast.

Output shape

a single JSON object containing the forecast results, model parameters, metrics, summary, limitations, and audit trail information.

Output includes forecast points with step indexes, the linear trend model parameters (slope, intercept, R-squared), error metrics (MAE, MSE, RMSE, MAPE), summary text of the forecast, warnings for input data quality and horizon choices, and detailed audit trail events. Requires an array of JSON objects as input, each representing one time series row with the specified value column containing numeric data. Produces error output if validation fails or insufficient observations exist.

Examples

Example 1: Forecasting a 3-step linear trend for a value column "closingPrice" with optional currency metadata.

Set Operation to 'Trend Forecast', specify 'closingPrice' as the Value Column, set Forecast Horizon to 3 (or desired number), optionally provide a Currency code, and optionally adjust Max Recommended Horizon (default 24). Input should be an array of objects representing the time series ordered sequentially.

Links

Discussion