1 views collected around this technical thread.
The JD Supply Chain algorithm team introduces TimeHF, a billion‑parameter time‑series large model that leverages RLHF to boost demand‑forecast accuracy by over 10%, detailing dataset construction, the PCTLM architecture, a custom RLHF framework (TPO), and extensive SOTA experimental results.
This article presents a pioneering billion‑parameter pure time‑series large model (PCTLM) trained on a 1.5‑billion‑sample dataset, introduces a novel RLHF framework (TPO) for time‑series forecasting, and demonstrates state‑of‑the‑art performance across multiple public benchmarks, surpassing existing models such as GPT4TS.
Pyraformer introduces a pyramidal attention mechanism that captures long-range dependencies in time-series data with linear time and space complexity, achieving state-of-the-art forecasting accuracy on multiple real-world datasets while reducing computational cost, as demonstrated in extensive ICLR-2022 experiments.
This article explains the Grey Prediction Model (GPM), especially the GM(1,1) variant, detailing its Accumulated Generating Operation, difference‑equation modeling, parameter estimation, and inverse transformation, and shows why it excels with limited, noisy data.
This article presents the SCNN model for multivariate time series forecasting, explains its decomposition into long‑term, seasonal, short‑term, and co‑evolving components, details the neural‑network‑based fusion and loss design, provides Python code snippets, and demonstrates its practical deployment for business volume prediction at Ctrip.
This article introduces the Grey‑Markov hybrid model, explains its theoretical foundations, outlines step‑by‑step modeling procedures, and demonstrates its superior forecasting performance on a consumer price index (CPI) case study, achieving a significant reduction in prediction error.
CausalMMM introduces an encoder‑decoder framework that automatically discovers heterogeneous, interpretable causal graphs among advertising channels while modeling temporal decay and saturation, using Granger‑based variational inference, and achieves over 5.7% improvement in causal structure learning and significant GMV prediction gains on Alibaba’s data.
The article offers a thorough examination of Facebook’s Prophet forecasting model, detailing its additive decomposition of trend, seasonality, holidays and regressors, the underlying Bayesian inference via Stan, the full training‑and‑prediction pipeline, data‑normalization tricks, uncertainty estimation, and practical source‑code insights for e‑commerce applications.
This article presents a comprehensive study on forecasting critical business metrics such as traffic, order volume, and GMV using traditional, machine‑learning, and deep‑learning time‑series models, detailing feature engineering, model design, experimental comparison, online deployment, and monitoring strategies.
This article presents a comprehensive case study of NIO's Power swap‑station ecosystem, detailing the business context, key forecasting challenges, the evolution from classical statistical models to deep‑learning architectures with specialized embeddings, and the practical outcomes and future plans for improving prediction accuracy.
This article presents NIO's smart energy service platform, focusing on the NIO Power swap‑station business and detailing how time‑series forecasting is applied to predict demand, addressing complex seasonality, holiday drift, growth and competition, and describing the underlying machine‑learning and deep‑learning models and system architecture.
This article explains the fundamentals of Recurrent Neural Networks and Long Short‑Term Memory units, their gating mechanisms, and demonstrates a practical Python Keras example that predicts future PM2.5 concentrations using an LSTM model.
This article reviews the award‑winning 2022 MCM/ICM C‑problem papers that develop and compare adaptive grid, ARIMA, LSTM, Prophet, and XGBoost‑based models for daily gold and Bitcoin trading, analyzing profitability, risk, transaction‑cost sensitivity, and providing evidence of superior strategy performance.
The paper proposes clustering thousands of related time series and training separate global LSTM models for each cluster, showing that this reduces heterogeneity, leverages shared information, and improves forecasting accuracy compared to individual models, with extensive experiments on CIF2016 and NN5 datasets.
This article surveys the most influential Transformer‑based research papers—from the original Attention Is All You Need work to recent models such as Autoformer and FEDformer—covering breakthroughs in natural language processing, computer vision, speech recognition, and long‑term series forecasting, and provides download links for each.
STARDOM is an end‑to‑end deep hierarchical forecasting model that jointly learns hierarchical constraints, query semantics via pretrained BERT, and a calibration matrix within an encoder‑decoder architecture, using a distilled reconciliation loss and hierarchical sampling to accurately predict large‑scale search traffic and outperform state‑of‑the‑art baselines.
This article presents a pandemic‑aware method for predicting national hotel monthly room‑nights over the next six months, detailing data augmentation, feature engineering, LSTM and SARIMA‑LASSO modeling, scenario‑based risk assessment, and evaluation results that demonstrate accurate forecasts despite COVID‑19 disruptions.
This tutorial explains how to build, train, and evaluate a multivariate LSTM model using Keras for hourly air‑pollution forecasting, covering data preparation, model design, prediction, and inverse scaling back to original units.
This article explains the structure and operation of recurrent neural networks (RNNs), their limitations, how long short‑term memory (LSTM) networks overcome these issues with gated mechanisms, and provides a complete Python implementation for time‑series airline passenger forecasting.
This article introduces the GM(2,1) grey model, presents its definitions and theorem, walks through a step‑by‑step case study with data preparation, parameter estimation, solution of the differential equation, and shows how to implement the whole process in Python with code examples and error analysis.