← Back to all news
arXiv Statistics ML By Sofia Gulevskaia, Mikhail Trapeznikov, Aleksandr Poslavsky, Alexander D'yakonov

Hierarchical Exponential-Gaussian Mixtures for Watch-Time Distribution Prediction

Read the original on arXiv Statistics ML →

The paper introduces Hierarchical Exponential-Gaussian Mixture (HEGM) models to predict watch-time distributions for short-video recommendations, addressing issues such as variance collapse, component redundancy, and inactive components found in the prior Exponential-Gaussian Mixture Network (EGMN). HEGM employs a hierarchical skip-watch decomposition, KL-based variance regularization, structured initialization, and removes certain regularizers to improve mixture stability and interpretability. Experiments on public and large-scale industrial datasets show that HEGM enhances ranking accuracy and threshold-event prediction while maintaining competitive point-estimation accuracy, and a 1.5‑month production A/B test demonstrates statistically significant engagement lifts.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Statistics ML.

Related stories

arXiv Machine Learning
Sep 25

RQ-Reg: A Residual-Quantization-Based Framework for Continuous Value Prediction in Recommender Systems

The paper introduces RQ-Reg, a residual‑quantization framework for predicting continuous values in recommender systems. It decomposes target values into a sequence of quantization codes, autoregressively refining predictions from coarse to fine granularity, and incorporates an ordinal‑aware objective to align embeddings with target order. Experiments on watch‑time, LTV, and a large‑scale online A/B test for GMV demonstrate competitive performance and strong generalization across diverse prediction tasks.

By Runpeng Cui, Zhipeng Sun, Chi Lu, Peng Jiang
ragefficiencybenchmarks
More like this →
arXiv Machine Learning
Sep 18

Fast Training of Mixture-of-Experts for Time Series Forecasting via Expert Loss Integration

The paper introduces an adaptive Mixture-of-Experts (MoE) framework for time series forecasting that incorporates expert-specific losses to give each expert a direct learning signal independent of gating weights. The overall objective combines base forecasting loss with these expert losses, encouraging experts to specialize on different temporal segments. A partial online learning strategy is added for efficient incremental updates, and experiments on economic, tourism, and energy datasets show the method outperforms state‑of‑the‑art neural models and foundation models, with ablation studies confirming the benefit of expert loss integration.

By Btissame El Mahtout, Florian Ziel
llmsbenchmarks
More like this →
arXiv Machine Learning
Sep 29

Aurora-X: Built for Extreme Time Series Forecasting

Aurora‑X is a billion‑parameter time‑series foundation model designed for extreme forecasting tasks. It employs a progressive curriculum that starts with channel‑independent pretraining, then adds cross‑variable dependencies, variable context and horizon lengths, and optional future covariates during mid‑training. A variable‑resolution post‑training stage allows adjustable temporal spans per token at inference, while a pattern‑guided mixture‑of‑experts expands capacity through sparse activation and expert specialization. An implicit quantile network head predicts arbitrary quantiles, enhancing probabilistic forecasting flexibility. Experiments on GIFT‑Eval, TIME, FEV‑Bench, TFB, and DAG‑Bench show state‑of‑the‑art performance against both pretrained TSFMs and task‑specific supervised models.

By Xingjian Wu, Chenjuan Guo, Xiangfei Qiu, Zhigang Hu, Hanyin Cheng, Peng Chen, Yang Shu, Jilin Hu, Bin Yang
benchmarks
More like this →