Initial commit: BAF Lakehouse fraud detection pipeline

End-to-end LightGBM fraud detection pipeline built as an R package,
orchestrated by targets with data stored in MinIO via Apache Arrow.
Includes 6-layer Lakehouse architecture, class imbalance tournament,
formally tuned hyperparameters (PR-AUC 0.198), and Quarto RevealJS slides.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

man/train_production_model.Rd

@@ -0,0 +1,30 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/functions.R
+\name{train_production_model}
+\alias{train_production_model}
+\title{Train and Serialize Production LightGBM Model}
+\usage{
+train_production_model(
+  data,
+  recipe,
+  best_params,
+  model_filename = "lgbm_prod.txt"
+)
+}
+\arguments{
+\item{data}{A data frame containing the full BAF dataset (Months 0-7).}
+
+\item{recipe}{A prepared tidymodels recipe.}
+
+\item{best_params}{A list or tibble of the winning hyperparameters.}
+
+\item{model_filename}{Character. The target filename. Defaults to "lgbm_prod.txt".}
+}
+\value{
+Character. The MinIO URI of the uploaded model artifact.
+}
+\description{
+Trains a LightGBM model on the complete dataset using the winning
+hyperparameters, serializes it to a text file, and uploads it directly
+to MinIO via the Apache Arrow S3 interface.
+}