Research map / Time series / Causal and generative time series
On the benefits of maximum likelihood estimation for Regression and Forecasting
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the ZNBP mixture likelihood with metric-specific inference instead of direct target-metric training | Regression and forecasting models · Favorita test | MAPE: 0.3199±0.0011 -> 0.3139±0.0011 | Helped |
| Use the ZNBP mixture likelihood with metric-specific inference instead of direct target-metric training | Regression and forecasting models · Favorita test | RMSE: 161.4919±0.4748 -> 164.6521±1.5185 | Hurt |
| Use the ZNBP mixture likelihood with metric-specific inference instead of direct target-metric training | Regression and forecasting models · M5 test | WAPE: 0.266±0.0001 -> 0.2677±0.0002 | Hurt |
| Use the ZNBP mixture likelihood with metric-specific inference instead of direct target-metric training | Regression and forecasting models · Bicycle Share test | RMSE: 878.5815±1.3059 -> 767.4368±7.1274 | Helped |
| Use the ZNBP mixture likelihood with metric-specific inference instead of direct target-metric training | Regression and forecasting models · Gas Turbine test | MAPE: 0.774±0.0054 -> 0.9877±0.0019 | Hurt |
| Use the ZNBP mixture likelihood and infer quantiles from its predicted distribution instead of quantile regression | Bicycle Share model · Bicycle Share test | p10QL: 0.0973±0.0002 -> 0.0788±0.0008 | Helped |
| Use the ZNBP mixture likelihood and infer quantiles from its predicted distribution instead of quantile regression | Bicycle Share model · Bicycle Share test | p90QL: 0.0628±0.0019 -> 0.0536±0.0007 | Helped |
| Add zero-inflation to Negative Binomial (ablation) | MLE mixture likelihood · Favorita test | MAPE: 0.3314+/-0.0016 -> 0.3186+/-0.0011 | Helped |
| Add zero-inflation to Negative Binomial (ablation) | MLE mixture likelihood · Favorita test | WAPE: 0.2521+/-0.002 -> 0.2453+/-0.002 | Helped |
| Add zero-inflation to Negative Binomial (ablation) | MLE mixture likelihood · Favorita test | RMSE: 175.501+/-1.1928 -> 170.0075+/-1.282 | Helped |
| Add Pareto to Zero-Inflated Negative Binomial (ablation) | MLE mixture likelihood · Favorita test | MAPE: 0.3186+/-0.0011 -> 0.3139±0.0011 | Helped |
| Add Pareto to Zero-Inflated Negative Binomial (ablation) | MLE mixture likelihood · Favorita test | WAPE: 0.2453+/-0.002 -> 0.2238±0.0009 | Helped |
Closest papers by meaning
- Maximum Likelihood Estimation is All You Need for Well-Specified Covariate Shift (ICLR 2024 · similarity 0.87)
- An Analysis of Linear Time Series Forecasting Models (ICML 2024 · similarity 0.84)
- The Bias-Variance Tradeoff in Data-Driven Optimization: A Local Misspecification Perspective (NeurIPS 2025 · similarity 0.84)
- DiffPTS: Rethinking Diffusion ELBO for Probabilistic Time Series Forecasting (NeurIPS 2026 · similarity 0.84)
- Leaving the Nest: Going beyond Local Loss Functions for Predict-Then-Optimize (AAAI 2024 · similarity 0.84)
- Risk Estimation in a Markov Cost Process: Lower and Upper Bounds (ICML 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.