Publications by T. Moudiki
tisthemachinelearner: A Lightweight interface to scikit-learn with 2 classes, Classifier and Regressor (in Python and R)
Contents Contents 1 – Python version 2 – R version 1 – Python version !pip install tisthemachinelearner import numpy as np from sklearn.datasets import load_diabetes, load_breast_cancer from sklearn.model_selection import train_test_split from tisthemachinelearner import Classifier, Regressor # Classification X, y = load_breast_cancer(ret...
494 sym R (9433 sym/2 pcs) 2 img
R version of survivalist: Probabilistic model-agnostic survival analysis using scikit-learn, xgboost, lightgbm (and conformal prediction)
This post is to be read in conjunction with https://thierrymoudiki.github.io/blog/2025/02/10/python/Benchmark-QRT-Cube and https://thierrymoudiki.github.io/blog/2024/12/15/python/agnostic-survival-analysis. Survival analysis is a group of Statistical/Machine Learning (ML) methods for predicting the time until an event of interest occurs. Examples o...
1418 sym 2 img
A simple test of the martingale hypothesis in esgtoolkit
Disclaimer: This is a work in progress. R code at the end. Introduction A fundamental utility in financial econometrics and stochastic processes is the martingale property, which implies that the best approximation of the future value of a time series or stochastic process, based on its historical values, is its present value. This property is crit...
5007 sym 2 img
Command Line Interface (CLI) for techtonique.net’s API
There is a new Command Line Interface (CLI) for Techtonique’s API, working on all operating systems (Windows, MacOS, Linux). For more details, on how to install it, see https://github.com/Techtonique/cli. Here are a few examples (there are more in the README). Export results to a JSON file: techtonique forecasting univariate /Users/t/Documents/da...
5227 sym Python (2461 sym/5 pcs) 8 img
Gradient-Boosting and Boostrap aggregating anything (alert: high performance): Part5, easier install and Rust backend
I recently released the genbooster Python package (usable from R), a Gradient-Boosting and Boostrap aggregating implementations that uses a Rust backend. Any base learner in the ensembles use randomized features as a form of feature engineering. The package was downloaded 3000 times in 5 days, so I guess it’s somehow useful. The last version of m...
1516 sym 2 img
Just got a paper on conformal prediction REJECTED by International Journal of Forecasting despite evidence on 30,000 time series (and more). What’s going on? Part2: 1311 time series from the Tourism competition
In #182, I presented a benchmark based on 30,000 time series, comparing the sequential split conformal method to the state of the art in Conformal prediction for Forecast. In this post, I’ll present a benchmark based on 1311 time series from the Tourism competition, comparing the sequential split conformal method to the state of the art. Theta m...
1827 sym R (8774 sym/4 pcs) 4 img
Unified interface and conformal prediction (calibrated prediction intervals) for R package forecast (and affiliates)
In the popular R package forecast, there are 2 different types of interfaces: A direct interface for functions like forecast::thetaf doing fitting and inference simultaneously nile.fcast <- forecast::thetaf(Nile) plot(nile.fcast) An interface for fitting first and then forecasting, like forecast::ets, where you need to use, in addition forecast...
1465 sym R (2962 sym/8 pcs) 14 img
GLMNet in Python: Generalized Linear Models
During the past few weeks, I’ve been adapting a Python version of the (seemingly abandoned?) official Stanford GLMNet package. Don’t try to build a programming interface on it yet, as it’s still “moving”. GLMNet implements the entire lasso or elastic-net regularization path for linear regression, logistic and multinomial regression models...
1240 sym Python (87828 sym/4 pcs) 2 img
Gradient-Boosting anything (alert: high performance): Part4, Time series forecasting
A few weeks ago, I introduced a model-agnostic gradient boosting (XGBoost, LightGBM, CatBoost-like) procedure for supervised regression and classification, that can use any base learner (available in R and Python package mlsauce). You can find the previous posts here: https://thierrymoudiki.github.io/blog/2024/10/06/python/r/genericboosting https:...
2139 sym Python (6161 sym/3 pcs) 4 img
Predictive scenarios simulation in R, Python and Excel using Techtonique API
In this post, we will learn how to simulate predictive scenarios using R, Python, and Excel, by using Techtonique API, available at https://www.techtonique.net. Input csv files used in the examples are available in Techtonique/datasets.The Excel version can be found in the Excel file https://github.com/thierrymoudiki/techtonique-excel-server/VBA-We...
4618 sym Python (1851 sym/2 pcs) 3 img