Tag:machine-learning
All the articles with the tag "machine-learning".
Tabular Foundation Models, Explained and Tested on a Real Sales Forecast
My notes on tabular foundation models (TabPFN, TabICL, TabFM): architecture, training, inference, licences, how they compare with trees, time series, and two tests against the LightGBM model we use for weekly store sales.
VN2 Features
The VN2 features I implemented and backtested in our LightGBM store forecast: definitions, correlation, SHAP analysis, the final feature set and how it compares with production, after a short background on the VN2 competition and its winning solution.
How a tabular foundation model predicts, step by step
One small churn table followed through a TabICL-style tabular foundation model: the three attention stages, how it is trained, what inference costs, how it compares with the original transformer and LLMs, and how to prepare your data.
Perfect Precision, Zero Recall, and No Curve to Trade Along
A document check cleared every forgery in a labelled set. The threshold was not the problem: the check read the file's structure while the forgeries were in the pixels. A seeded rebuild shows a structure-only detector reaching exactly two operating points, AUC 0.500, and what calibrating on genuine documents alone can and cannot buy.
The Constant Nobody Could Re-Measure
A magic number survives because there is nothing to re-measure it against. The calibration set is usually already in a log you throw away, the same data sets the noise floor that makes any result readable, and the estimator you fix is rarely where the error was.
Calendar Features for TabPFN Time Series: Geometry, Encodings, and an Open Edge Case
TabPFN-TS turns forecasting into a table of features. A 24-hour edge case shows that a correct encoding, compatibility with the trained model, and evidence for a release are three separate questions.
Forecast Metrics
My notes on forecast metrics for weekly sales forecasts: MAE, bias, MAPE, WAPE, MASE and RMSSE, business KPIs from a split bias, censored demand in backtests, and why RelMAE and RelMSE replaced MASE and RMSSE.
Solving the Low and High Seller Bias in a Weekly Sales Forecast
Two studies on a LightGBM weekly sales forecast that under-forecasts high sellers and over-forecasts low sellers: separate models, a better initialisation, and normalizing the dataset, then a wide backtest of normalization.
Validation, Hyperparameter and Feature Set Studies for a Weekly Sales Forecast
Three studies on a LightGBM weekly sales forecast, in the order we ran them: the validation set for early stopping, the hyperparameter set, and the feature set.
Explainability Using SHAP
How I used SHAP to explain a LightGBM sales forecast: the plots, an example of an analysis, feature correlation with SHAP, a template notebook and an explainability dashboard.