MLFS

Machine Learning Forest Simulator (MLFS) is a data-driven forest development model implemented as an R package. MLFS aims to remove the need for hand-tuned model parametrization and provide an easy-to-use, freely available tool applicable across forest types—from even-aged monocultures to mixed, vertically structured stands.

MLFS is age- and spatially independent. It uses repeated forest inventory data (at least two periods) to:

1. model basal area increment (BAI),
2. update tree and crown heights at each simulation step,
3. simulate individual-tree mortality,
4. simulate ingrowth of new trees, and
5. represent harvesting.

The main inputs are forest inventory plot data and site descriptors, which are used to train modular sub-models.

Key features

• Data-driven: learns directly from repeated inventory data.
• Modular workflow: separate sub-models for growth, height dynamics, mortality, ingrowth, and harvesting.
• Broad applicability: works from even-aged to mixed, multi-layered forests.
• R-native: integrates with common R tooling for analysis and reproducibility.

Installation

Install the released version from CRAN:

install.packages("MLFS")

Or install the development version from GitHub:

# install.packages("remotes")
# remotes::install_github("jernejjevsenak/MLFS")
install.packages("MLFS", repos = c('https://jernejjevsenak.r-universe.dev', 'https://cloud.r-project.org'))

Input data requirements (at a glance)

• Forest inventory plots with at least two measurement periods.
• Tree-level attributes (e.g., species, DBH; optional heights/crown metrics if available).
• Site descriptors used for training sub-models.

Getting started

After installation:

library(MLFS)

# Explore package documentation
help(package = "MLFS")

A typical workflow is:

• Prepare and validate inventory and site tables.
• Train sub-models using repeated measurements.
• Run simulations over chosen time steps.
• Summarize and visualize outputs at tree/plot/stand levels.

Help & issues

• Report bugs or request features via GitHub Issues: https://github.com/jernejjevsenak/MLFS/issues

How to cite

citation("MLFS")

Authors

Jernej Jevšenak