> ## Documentation Index
> Fetch the complete documentation index at: https://docs.fincept.in/llms.txt
> Use this file to discover all available pages before exploring further.

# Model update and transfer

> Reuse a trained forecaster: append new observations without retraining and forecast from the new end; forecast series the models never saw (a new ticker, store or market) with intervals recalibrated, rescaled or density-ratio reweighted for them; covariate-shift weights between two feature sets.

Toolset `forecast_transfer`: 3 tools.

| Tool | What it does | Notes |
| - | - | - |
| `forecast_update` | Trains on the history, scores that model on the new observations it never saw (an honest out-of-sample check), then appends them to the stored series without retraining and forecasts h periods from the new end with the original conformal intervals. | |
| `forecast_transfer` | Trains models on source series (data or y) and forecasts different target series with them (transfer learning: a new ticker, product or market with short history), with conformal intervals adapted to the targets by recalibration, error or scale ratios, or density-ratio reweighting. | |
| `forecast_density_ratio` | Density-ratio weights w(x) = p\_target(x) / p\_source(x) for every source row, from a classifier trained to tell target rows from source rows on standardized features: how representative each past observation is of the target conditions. | |

Inputs, limits and outputs are described in [Fincept Forecast](/guides/fincept-forecast). Full schemas: `fincept_describe_tool`.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.