Tagged articles
forecasting
57 articles · Page 1 of 1
The UniCM model unifies ocean‑atmosphere climate modes in a dual‑branch transformer, achieving record‑long ENSO forecasts and revealing emergent predictability across seven key global modes, while offering interpretable attention maps that turn AI from a pure predictor into a climate discovery tool.
AI for ScienceInterpretabilityTransformer
0 likes · 10 min read

This lesson explains how to set baseline reference models, use MAPE and MSE as quantitative evaluation metrics, and demonstrates their application on quarterly EPS data and simulated random walks to reveal the limits of predictability in non‑stationary series.
MSEMaPEPython
0 likes · 13 min read

The article reviews why classic classification augmentations fail for forecasting, outlines a taxonomy of effective time‑series augmentation techniques—including frequency‑domain, decomposition, and patch‑based methods—details the Temporal Patch Shuffle (TPS) pipeline, and presents extensive experiments showing TPS achieves state‑of‑the‑art improvements across long‑term, short‑term, and classification tasks.
Data AugmentationTemporal Patch ShuffleTime Series
0 likes · 17 min read

The article reviews the HORAI model, which introduces a frequency‑enhanced multimodal pretraining paradigm and the massive MM‑TS dataset, showing that integrating derived images, endogenous text, and real‑world news dramatically improves zero‑shot forecasting and anomaly detection across six domains.
HORAIPretrainingTime Series
0 likes · 23 min read

This article presents a comprehensive framework for forecasting that combines peer (same‑level) comparison, bottom‑up decomposition, top‑down system thinking, time‑series analysis, causal modeling, and scenario simulation, while highlighting each method's strengths, limitations, and practical wisdom for effective decision‑making.
Scenario AnalysisTime Seriescausality
0 likes · 12 min read

This article reviews key mathematical modeling techniques for renewable energy, covering solar photovoltaic irradiance and efficiency models, wind turbine power and Weibull wind speed distributions, battery storage dynamics and capacity optimization, microgrid energy management, and forecasting methods such as ARMA and MPC, highlighting their role in efficient clean energy deployment.
Renewable Energyenergy storageforecasting
0 likes · 10 min read

This digest summarizes three recent AI papers—DoFlow, Forecast2Anomaly, and ForecastGAN—detailing their causal generative flow model for interventions, a retrieval‑augmented framework for zero‑shot anomaly prediction, and a decomposition‑based adversarial approach that improves multi‑horizon forecasting across diverse datasets.
Causal InferenceTime Seriesanomaly detection
0 likes · 8 min read

This article introduces the core principles of Grey Relational Analysis and the Grey Prediction Model, explains their calculation steps, and demonstrates how they can be applied across engineering, economics, and environmental fields to analyze limited data, select key indicators, evaluate systems, and forecast trends.
Grey Prediction ModelGrey Theorydata modeling
0 likes · 8 min read

This article summarizes four recent time‑series forecasting papers, covering a universal delay‑embedding foundation model, a dual causal network that leverages exogenous variables, a distribution‑aware alignment plug‑in called TimeAlign, and a shapelet‑based framework for interpretable directional forecasting in noisy financial markets.
Representation LearningTime Seriescausal network
0 likes · 9 min read

The HGTS‑Former model introduces a hierarchical hypergraph backbone combined with a Transformer to capture high‑order and dynamic dependencies in multivariate time‑series data, and experimental results on eight datasets show it consistently outperforms state‑of‑the‑art methods in both long‑term forecasting and interpolation tasks.
HGTS-FormerHypergraphTransformer
0 likes · 11 min read

It presents concise summaries of six recent arXiv papers on unsupervised domain adaptation, efficient forecasting, SHAP explanations, text‑reinforced multimodal forecasting, online prediction with feature adjustment, zero‑shot forecasting zoo, and a new anomaly‑detection metric, highlighting methods, datasets, and results.
SHAPTime Seriesanomaly detection
0 likes · 16 min read

The article summarizes Professor Long Mingsheng’s presentation on the Timer series of time‑series large models, detailing the three core challenges of industrial time‑series analysis, the evolution from statistical methods to generative models, and the technical breakthroughs of Timer 1.0, 2.0 and 3.0 that enable multi‑task, long‑context, and trillion‑scale forecasting for industrial digital transformation.
Generative AIIoTDBforecasting
0 likes · 14 min read

The article explains how relying solely on historical data for prediction often leads to large errors because future structural changes and missing variables are ignored, and it proposes causal modeling, scenario simulation, and real‑time signals as more reliable forecasting approaches.
Causal Modelingbig dataforecasting
0 likes · 9 min read

Traditional fixed‑threshold monitoring often triggers noisy alerts during routine business rhythms, but by modeling time‑series patterns with Facebook Prophet to predict dynamic confidence intervals, teams can automatically adjust thresholds, reduce false positives, and accurately detect true anomalies across diverse services.
ProphetTime Seriesanomaly detection
0 likes · 7 min read

This article dissects common failures in e‑commerce replenishment—such as hot‑product black holes, slow‑moving stock graves, and supply‑chain avalanches—and presents a seven‑step framework of dynamic forecasting, tiered strategies, distributed inventory, and automated safeguards to stabilize inventory levels.
Operationsautomationforecasting
0 likes · 9 min read

Using a year‑long revenue dataset from an online‑education firm, this article examines how description, causal explanation, and statistical modeling together reveal patterns, uncover underlying drivers, and highlight the limits and uncertainties of forecasting future performance.
Business AnalyticsCausal Inferenceforecasting
0 likes · 7 min read

This article reviews the characteristics of financial time series, outlines a standard forecasting workflow, and compares traditional statistical models (ARMA, SARIMAX) with recent deep‑learning approaches (TimesNet, TimeMixer, PatchTST), highlighting their strengths, limitations, and typical use cases in finance.
ARMAPatchTSTSARIMAX
0 likes · 19 min read

The article explains how modern supply chain forecasting has shifted from qualitative expert judgment to quantitative AI-driven methods such as DeepAR, ensemble learning, and Transformers, and outlines the skills needed for practitioners to build effective predictive models.
AIDeepARTransformer
0 likes · 10 min read

This article explains how to extract a variety of date‑time based features—including date, time, lag, rolling, expanding, and domain‑specific attributes—from a time‑series dataset using pandas, and discusses proper validation strategies for building reliable forecasting models.
PandasTime Seriesfeature engineering
0 likes · 14 min read

This article presents Ant Group's comprehensive time‑series AI solutions, detailing the AntFlux platform, the evolution from statistical to deep and large‑scale models—including Time‑LLM, iTransformer, and SLOTH—and illustrating how these technologies empower business insight, forecasting, decision‑making, and green computing across diverse scenarios.
AntFluxTime Seriesforecasting
0 likes · 17 min read

Alibaba Cloud’s big‑data research team announced four groundbreaking time‑series papers accepted at ICLR 2024, ICDE 2024 and WWW 2024, introducing models such as Pathformer, ContraLSP, MACE, and LARA that advance multi‑scale forecasting, explainable AI, and efficient anomaly detection for intelligent operations.
AITime Seriesanomaly detection
0 likes · 11 min read

Grey Theory, introduced by Deng Julong in 1982, offers a mathematical framework for analyzing systems with incomplete or uncertain data, using techniques like generated series and the GM(1,1) model to enable reliable forecasting and decision‑making across fields such as economics, environment, and product lifecycle analysis.
Grey TheoryLimited Datadata modeling
0 likes · 8 min read

Ant Group shares its time‑series AI practice, detailing the AntFlux intelligent engine, the evolution of statistical and deep learning models, large‑scale time‑series platforms, and real‑world applications across finance, cloud, and green computing, illustrating challenges, innovations, and future directions.
AntFluxforecastingindustrial AI
0 likes · 19 min read

The article details JD’s smart supply‑chain innovations, describing an explainable AI forecasting method that boosts prediction accuracy while maintaining interpretability, and an end‑to‑end inventory management model based on multi‑quantile RNNs that improves replenishment decisions, reduces costs, and enhances overall operational efficiency.
explainable AIforecastinginventory management
0 likes · 14 min read

This article presents Ant Group's comprehensive time‑series AI solutions, covering the business value of temporal data, the evolution of statistical and deep learning models, large‑scale time‑series platforms such as AntFlux, and real‑world applications ranging from financial forecasting to green computing.
AIAntFluxTime Series
0 likes · 17 min read

Didi balances ride‑hailing supply and demand by forecasting regional needs with time‑series and deep‑learning models, then optimally repositioning drivers through integer programming and refining policies via imitation and offline reinforcement learning, ultimately enhancing passenger experience and platform efficiency.
DidiOffline Reinforcement Learningforecasting
0 likes · 16 min read

The article presents a practical toolbox of statistical and machine‑learning metrics—including short‑term growth rates, CAGR, Excel forecasting functions, Wilson score adjustment, sigmoid decay weighting, correlation coefficients, KL divergence, elbow detection with KneeLocator, entropy‑based weighting, PCA, and TF‑IDF—offering concise formulas and code snippets for data analysis without deep theory.
PCAcorrelationdata analysis
0 likes · 12 min read

This article presents JD's super automation approach for its supply chain, detailing the business background, challenges, AI‑driven forecasting, procurement, intelligent allocation, inventory clearing, integrated decision making, and future directions toward fully automated, optimal end‑to‑end operations.
JD.comforecastingmachine learning
0 likes · 17 min read

This presentation outlines Alibaba's digital supply chain strategy, detailing the data, analysis, and decision challenges, the multi‑layer digitalization and intelligent solutions, the evolution of the Falcon forecasting technology, and the Alibaba DChain Forecast SaaS product, with case studies and a Q&A.
AIAlibabaDigitalization
0 likes · 22 min read

This article explains the Leslie matrix model for age‑structured population forecasting, outlines its mathematical formulation, demonstrates how to build and solve it using Python code, and shows how to derive demographic indicators such as average age, lifespan, aging index, and dependency ratio.
Leslie matrixPythonage‑structured model
0 likes · 9 min read

This article introduces the Holt‑Winters additive exponential smoothing model, explains its mathematical formulation and when to use additive versus multiplicative versions, and provides Python examples using statsmodels to fit both exponential and linear trend variations, illustrated with plots.
Holt-WintersPythonexponential smoothing
0 likes · 5 min read

This article explains the fundamentals of weakly stationary time series, defines mean, variance, autocovariance, and autocorrelation functions, introduces AR, MA, ARMA, and ARIMA models, discusses model identification using ACF/PACF, selection criteria like AIC/SBC, diagnostic testing, and provides Python statsmodels code examples for implementation.
ARMAPythonforecasting
0 likes · 18 min read

This article explains the principles, formulas, and Python implementations of simple and double moving averages, as well as single, double, and Holt's exponential smoothing methods, illustrating each technique with real data cases and code samples.
exponential smoothingforecastingmoving average
0 likes · 16 min read

This tutorial explains how to convert single‑variable and multivariate time‑series data into supervised learning formats using pandas shift() and a custom series_to_supervised() function, covering one‑step, multi‑step, and sequence forecasting examples with complete Python code.
PythonTime Seriesforecasting
0 likes · 18 min read

This article explains how to identify, order‑select, estimate parameters, validate, and forecast ARMA time‑series models, covering the Akaike Information Criterion, various estimation techniques, and diagnostic tests such as the Ljung‑Box test, with practical Python implementation guidance.
AICARMAParameter Estimation
0 likes · 4 min read

This guide explains the limitations of ARIMA‑type models for handling changing variance, introduces ARCH and GARCH as solutions, and walks through Python implementations—including data generation, model fitting, and forecasting—complete with code snippets and visualizations.
ArchGARCHPython
0 likes · 7 min read

A time series is a chronologically ordered set of interrelated data points whose analysis involves studying its development patterns and forecasting future behavior, with classifications based on dimensionality, continuity, statistical properties such as stationarity, and distribution types like Gaussian or non‑Gaussian.
Time Seriesforecastingmultivariate
0 likes · 2 min read

This guide explains the theory behind Vector Autoregression (VAR) models, outlines the complete modeling workflow—including data preparation, stationarity and cointegration testing, lag order selection, parameter estimation, stability diagnostics, and impulse‑response and variance‑decomposition analysis—and provides a full Python implementation with code examples.
PythonTime Serieseconometrics
0 likes · 9 min read

This article explains the recursive forecasting formulas for ARMA and MA(q) time‑series models, showing how forecasts depend only on past observations, how model invertibility ensures stability, and how estimated parameters are used in practical prediction.
ARMAMA(q)Time Series
0 likes · 2 min read

This article explains the process of constructing ARMA models, covering model identification, order selection using the AIC criterion, parameter estimation (including Python implementation), and diagnostic testing such as Ljung‑Box, before demonstrating how to generate forecasts from the fitted model.
AICARMATime Series
0 likes · 4 min read

This article introduces Grey System Theory, explains its origins, core concepts of partial information, advantages over black‑box models, data accumulation/reduction techniques, and demonstrates a Python case study that improves forecasting accuracy for short‑term exponential trends.
data-accumulationforecastinggrey-system
0 likes · 8 min read

This tutorial explains how to convert single‑variable and multi‑variable time‑series data into a supervised‑learning format using Pandas' shift() function and a custom series_to_supervised() helper, covering one‑step, multi‑step, and multivariate forecasting examples with complete Python code.
PandasPythonTime Series
0 likes · 20 min read

This article details Alibaba's exploration of intelligent supply‑chain forecasting, covering scenario classification, three generations of prediction algorithms, the self‑developed Falcon model, performance evaluation, and real‑world cases such as Double 11 and live‑streaming, highlighting challenges and practical solutions.
AITime Seriesdeep learning
0 likes · 18 min read

This article explains a NeurIPS 2021 paper that introduces a learnable autocorrelation correction for neural network time‑series models, shows the underlying theory, provides concise PyTorch code implementing the adjustment, reports a ~17% average performance gain across datasets, and lists additional practical tricks for time‑series forecasting.
PyTorchTime Seriesautocorrelation
0 likes · 6 min read

This article outlines Ctrip's approach to time series forecasting, covering background, common tools such as factor‑based models, traditional statistical methods like ARIMA, and machine‑learning techniques including tree and neural networks, and shares practical experiences on data splitting, feature engineering, model stability, and evaluation.
ARIMACtripTime Series
0 likes · 13 min read

JD.com's intelligent supply‑chain team combined AI‑driven forecasting, S&OP planning, real‑time inventory response, smart fulfillment, anti‑arbitrage detection, price governance, and precise C2M delivery to dramatically cut costs, improve inventory turnover, and deliver a seamless 11.11 shopping experience.
OperationsPrice Optimizationartificial intelligence
0 likes · 18 min read

This article introduces time series fundamentals, classification, and challenges for internet businesses, then provides a step‑by‑step Python tutorial on ARIMA modeling—including data loading, stationarity testing, differencing, ACF/PACF analysis, AIC‑based order selection, model training, prediction, error evaluation, exogenous variable integration, and diagnostic checks.
ARIMAPythonTime Series
0 likes · 11 min read

This article provides a comprehensive tutorial on Prophet's key parameters, their meanings, and practical tips for tuning them—including growth, changepoints, seasonalities, holidays, and Bayesian settings—along with Python code examples for grid search and cross‑validation to improve forecasting accuracy.
ProphetPythonTime Series
0 likes · 14 min read

The article explains how advertising inventory forecasting—predicting how many users will see a specific ad—poses challenges due to fluctuating traffic and user segmentation, and describes a new deep spatial‑temporal tensor factorization model that dramatically improves prediction accuracy, scalability, and robustness for large‑scale ad platforms.
AIAdvertisingTime Series
0 likes · 11 min read

The article presents a complete workflow for predicting next‑day API request volumes by exploring per‑minute traffic data, handling missing values, applying seasonal decomposition, training an ARIMA model on the trend component, and generating confidence intervals to flag anomalous spikes.
ARIMATime Seriesanomaly detection
0 likes · 12 min read

The article reviews traditional, neural network, and open‑source time‑series forecasting methods, explains their strengths and limitations, and demonstrates how NetEase applies short‑term and long‑term prediction models such as Holt‑Winters, ARIMA, STL, Prophet, and LSTM to improve game monitoring and proactive alerting.
AIOpsARIMAHolt-Winters
0 likes · 12 min read

JD's chief scientist Shen Zuo‑jun explains how shortening the decision chain with end‑to‑end algorithms and intelligent multi‑level replenishment dramatically improves inventory turnover, stock availability, and forecasting accuracy, showcasing a novel supply‑chain research direction that integrates AI, big data, and human expertise.
End-to-EndOperationsforecasting
0 likes · 9 min read


JD TechSep 29, 2018 · Artificial Intelligence
The article outlines JD.com's evolution of prediction technology from early book‑category sales forecasting to a comprehensive AI‑driven platform that supports sales, order, and GMV forecasts, describes its modular architecture and core algorithm choices, and discusses future enhancements for smarter supply‑chain collaboration.
big dataforecastingmachine learning
0 likes · 6 min read

This article describes Ctrip's call‑center staffing challenge, detailing data cleaning, trend analysis, feature engineering, the initial ARIMAX‑Fourier model (V1.0), its limitations, and the improved V2.0 solution that combines TBATS, ARIMA residuals and XGBoost, achieving up to 89.5% prediction accuracy.
Time SeriesXGBoostcall center
0 likes · 9 min read

This article explains how to process a periodic time‑series, split it into training and test sets, smooth the data, decompose it with statsmodels' seasonal_decompose, forecast the trend using an ARIMA model, and evaluate the results with RMSE, providing a practical workflow for accurate forecasting.
ARIMAPythonTime Series
0 likes · 5 min read

The article explains how a data analyst can use statistical and machine‑learning models such as linear regression, tree‑based boosting, STL decomposition, and Prophet for both non‑time‑series forecasting and time‑series monitoring, highlighting data‑quality concerns, feature‑engineering practices, and deployment considerations like PMML packaging.
BIProphetSTL
0 likes · 13 min read

This article explains how to estimate future daily active users (DAU) for an app by modeling the accumulation of new users and their retention decay, addressing challenges of changing historical retention rates and proposing a combined approach using recent averages and curve‑fitted functions to predict long‑term user activity.
DAUUser Acquisitionforecasting
0 likes · 9 min read
