
NGBoost: Natural Gradient Boosting for Probabilistic Prediction
We present Natural Gradient Boosting (NGBoost), an algorithm which bring...
read it

Learning Probabilistic Ordinal Embeddings for UncertaintyAware Regression
Uncertainty is the only certainty there is. Modeling data uncertainty is...
read it

Uncertainty in Gradient Boosting via Ensembles
Gradient boosting is a powerful machine learning technique that is parti...
read it

Probabilistic Gradient Boosting Machines for LargeScale Probabilistic Regression
Gradient Boosting Machines (GBM) are hugely popular for solving tabular ...
read it

Transformationbased generalized spatial regression using the spmoran package: Case study examples
This study presents application examples of generalized spatial regressi...
read it

Probabilistic Regression for Visual Tracking
Visual tracking is fundamentally the problem of regressing the state of ...
read it

Uncertaintyaware INVASE: Enhanced Breast Cancer Diagnosis Feature Selection
In this paper, we present an uncertaintyaware INVASE to quantify predic...
read it
Multivariate Probabilistic Regression with Natural Gradient Boosting
Many singletarget regression problems require estimates of uncertainty along with the point predictions. Probabilistic regression algorithms are wellsuited for these tasks. However, the options are much more limited when the prediction target is multivariate and a joint measure of uncertainty is required. For example, in predicting a 2D velocity vector a joint uncertainty would quantify the probability of any vector in the plane, which would be more expressive than two separate uncertainties on the x and y components. To enable joint probabilistic regression, we propose a Natural Gradient Boosting (NGBoost) approach based on nonparametrically modeling the conditional parameters of the multivariate predictive distribution. Our method is robust, works outofthebox without extensive tuning, is modular with respect to the assumed target distribution, and performs competitively in comparison to existing approaches. We demonstrate these claims in simulation and with a case study predicting twodimensional oceanographic velocity data. An implementation of our method is available at https://github.com/stanfordmlgroup/ngboost.
READ FULL TEXT
Comments
There are no comments yet.