If you are around and interested in: 🔹 Causal Inference 🔹 Time-series analysis 🔹 Probabilistic modeling / GPs

Please feel free to DM me or let’s grab a coffee during the breaks! I’d love to exchange ideas.

#MachineLearning #CausalInference #GaussianProcess #EnvironmentalData #Singapore #Conference

A framework using #GaussianProcess Regression + SHAP helps quantify how neurological + demographic factors impact walking performance in people with #SpinalCordInjury.

Read the study: ieeexplore.ieee.org/document/11215849

Relative Information Gain Improves Gaussian Process Regression

Researchers propose 'relative information gain', a metric that improves GP regression with tighter bounds. The 28‑page paper was submitted on 5 Oct 2025. Read more: getnews.me/relative-information-gai... #relativeinformationgain #gaussianprocess

Trained Shallow Neural Networks Converge to Gaussian Processes

Research presented in 2025 gives upper bounds on Wasserstein distance between trained shallow networks and their Gaussian process limit, showing polynomial decay with width. getnews.me/trained-shallow-neural-n... #neuralnetworks #gaussianprocess

Graph Random Features Enable Faster Gaussian Processes on Large Graphs

Graph random features (GRFs) reduce Gaussian-process cost from O(N³) to O(N^{3/2}), allowing inference on graphs with over 1 million nodes on a single chip. getnews.me/graph-random-features-en... #graphrandomfeatures #gaussianprocess

InqEduAgent Offers Adaptive AI Learning Partners via Gaussian Process

InqEduAgent pairs learners with AI‑driven virtual study partners and uses Gaussian‑process matching to adapt partner selection; the code and data are open on GitHub. Read more: getnews.me/inqeduagent-offers-adapt... #ai #gaussianprocess #education

Gaussian Process Adaptive Tracking Enhances Autonomous Vehicles

A new trajectory‑tracking method blends Gaussian Process learning, validated on a 1/10‑scale electric race‑car; the paper was first submitted Jan 2025. Read more: getnews.me/gaussian-process-adaptiv... #gaussianprocess #autonomousvehicles

New Method Calibrates Location Errors for More Accurate Radio Maps

A method calibrates location errors in Gaussian-process radio maps, keeping RMSE loss to about 0.25-0.29 dB for sensor positions. Tested on mobility data, it beats baselines by 1 dB. getnews.me/new-method-calibrates-lo... #radiomaps #gaussianprocess

Bayesian Algorithms for Adversarial Online Learning in Infinite Actions

Researchers propose a Thompson sampling that uses a Gaussian‑process prior on the d‑dimensional unit cube [0,1]^d and achieves a regret bound of O(β√(Td log(1+√d λ/β))). getnews.me/bayesian-algorithms-for-... #thompsonsampling #gaussianprocess

Gaussian Process Policy Iteration with Schwarz Acceleration for HJB

GP policy iteration with additive Schwarz acceleration speeds HJB and mean‑field game convergence; the arXiv preprint was posted 19 Sep 2025. Read more: getnews.me/gaussian-process-policy-... #gaussianprocess #hjb

New Metrics for Identifying Variables and Transients in Large Astronomical Surveys A key science goal of large sky surveys such as those conducted by the Vera C. Rubin Observatory and precursors to the Square Kilometre Array is the identification of variable and transient objects. O...

My manuscript has hit the #arXiv. Take a look at what I've been thinking about for the past couple of years. It's far from pioneering or perfect, but it's my first step into #astrostatistics. In other news, I have become #GaussianProcess proselytiser-in-chief. #firstpost arxiv.org/abs/2508.09441