‘Zero-Shot Probabilistic Stock Returns Forecasting with Pretrained RVFL Networks’ accepted at COPA 2026 (and to appear in the Proceedings of Machine Learning Research)
Link to the paper and the code repository.
Thierry Moudiki’s webpage
Personal webpage:
Latest posts
‘PCARVFLSimulator’: a GAN-like tabular data synthesizer built from PCA scores, a Random Vector Functional-Link network, and residuals bootstrapping
A lightweight, GAN-like tabular data synthesizer that swaps the adversarial training loop for PCA, a Random Vector Functional-Link network, and residuals bootstrapping.
‘garchf’: GARCH probabilistic forecasting with package ‘forecast’-style interface (and ‘rugarch’ under the hood)
The package garchf wraps the extensive GARCH modeling machinery in rugarch behind a forecast-style interface, so that GARCH-based forecasts can be produced, plotted, and evaluated with the same conventions used throughout the forecast ecosystem (in the spirit of functions like forecast::thetaf()). The workhorse function, xgarchf(), fits a GARCH model to a univariate time series – with a choice of…
GPopt for R: Bayesian and conformal optimization of black-box functions and hyperparameter tuning
GPopt for R: Bayesian and conformal optimization of black-box functions and hyperparameter tuning
My last R posts: How conformalization helps weak models, fast conformal prediction with jackknife+ (and no refitting), and sklearn in R
My last R posts: How conformalization helps weak models, fast conformal prediction with jackknife+ (and no refitting), and sklearn in R.
Natively Interpretable Boosting
This notebook works with a booster, cybooster, that is interpretable by construction instead of relying on post hoc methods like SHAP or LIME.
Fast conformal prediction (no refitting) for some Machine Learning models via closed-form jackknife plus
Fast conformal prediction for some Machine Learning models via closed-form jackknife plus (no refitting involved, just Linear Algebra).
Using scikit-learn models in R easily with the tisthemachinelearner package
Using the tisthemachinelearner R package to train scikit-learn models and compute prediction intervals
No-Code Machine Learning in Excel with the Techtonique API
Using the Techtonique API in Excel without writing any code
How Conformal Prediction Makes Linear Models Good Enough — An Example Using R Package mlS3
A comparison of split conformal prediction across several predictive models, using R package mlS3, on the BostonHousing dataset