The article uses two statistical models to explain why participants in the Henan “Three Supports, One Assistance” exam cheat to scores around 80 rather than the lower 70, and why a simple score‑threshold rule cannot effectively deter such cheating, offering concrete policy suggestions.
This article explains the fundamental differences between anomaly and novelty detection, outlines how to model univariate outliers using probability distribution fitting with the distfit library, and demonstrates the workflow on synthetic height data and real natural‑gas price data, including model selection, visualization, and prediction.
This article presents eight essential probability distributions for everyday data‑science tasks, explains when to use each, provides concise Python code for fitting and sampling, and shares practical tips and a real‑world case study to help you choose the right model quickly.
This article explores how power‑law and other heavy‑tailed distributions appear in content ecosystems, explains their statistical foundations, discusses why they are common, and presents data‑driven strategies—including integer programming, graph‑based creator analysis, and causal inference—to optimize content production, recommendation, and settlement policies.
Lognormal distribution, where a variable’s logarithm follows a normal law, offers non‑negative, right‑skewed modeling ideal for phenomena such as rainfall, river flow, asset prices, and biological sizes, and this article explains its definition, properties, and a practical rainfall‑modeling case study.
Linear regression often fails to capture nested data structures, but hierarchical (multilevel) linear models address this limitation by modeling both within‑group and between‑group variation, enabling nuanced analysis of factors like school type on student performance and extending to fields such as ecology and health.
The whitepaper released by Volcano Engine and AMD reveals how QoS variations affect multimedia QoE, introduces the MPEI model with S‑, J‑, and D‑type features, and provides a comprehensive metric system to help businesses quantify and optimize video experience for growth.
This article explains the difference between correlation and causation, illustrates common misconceptions with real‑world examples, and introduces statistical tools such as randomized experiments, instrumental variables, propensity score matching, and difference‑in‑differences that help researchers uncover genuine causal effects in mathematical modeling.
This article provides comprehensive regression analysis methods and Python code examples for various business scenarios including e-commerce, market research, healthcare, finance, social media, HR, education, hospitality, marketing, and logistics.
This article introduces the Python Statsmodels library, explains its key features such as linear regression, GLM, time‑series and robust methods, shows how to install it, load data with pandas, perform descriptive statistics, visualizations, hypothesis testing, and simple and multiple linear regression examples.
This article explores how data in content ecosystems is distributed, contrasting the classic normal distribution with heavy‑tailed power‑law patterns, explains why power‑law appears frequently, discusses its statistical properties and risks, and presents practical optimization and causal‑inference methods applied to creator incentives and platform strategies.
These COMAP MCM problem statements challenge teams to develop statistical models for pricing used sailboats using a large 2023 dataset and to propose innovative strategies for the Olympic Games, evaluating regional effects, data sources, and policy recommendations for sustainable hosting.
This article explains logistic regression for binary outcomes, demonstrates its implementation in R with the TitanicSurvival dataset, and interprets the model coefficients showing how gender, age, and passenger class significantly affect survival probability.
This article explains the theory of multiple linear regression, demonstrates how to fit such a model in R using the birthwt dataset with the lm() function, and interprets the output, diagnostic plots, and handling of categorical variables.
This article explains the fundamentals of multiple linear regression, covering model formulation, least‑squares estimation of coefficients, statistical tests for significance, and how to use the fitted equation for accurate predictions and confidence intervals.
This article explains how regression can handle grouped (aggregated) data using weighted least squares, illustrates the impact of heteroskedasticity, and shows how dummy variables encode categorical factors for flexible, non‑parametric modeling of treatment effects.
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.
This article explains the fundamentals of multiple linear regression, covering model formulation, least‑squares estimation of coefficients, hypothesis testing of the regression equation, and how to use the fitted model for point and interval predictions.
This article explains the theory and practice of simple linear regression, covering deterministic vs. stochastic relationships, the least‑squares estimation of coefficients, goodness‑of‑fit measures such as R², hypothesis testing for linearity, and a real‑world case linking wine consumption to heart‑disease mortality.
This article explains the fundamentals of multiple linear regression, covering model formulation, least‑squares estimation of coefficients, statistical tests for significance, and how to use the fitted equation for point and interval predictions.
The Tsinghua‑Meituan Digital Life Joint Research Institute’s academic salon will feature Associate Professor Jiang Rui presenting a cyclic generative adversarial network for probability density estimation, demonstrating how merging statistical models with deep‑learning techniques can solve core statistical problems and foster industry‑academia innovation.
This article introduces survival analysis, explains how to model user churn by defining purchase and cancellation times as birth and death events, describes data formatting, presents descriptive Kaplan‑Meier results, and shows how Cox regression quantifies the impact of factors such as membership and activity on user survival.
This tutorial walks through the full Python-based ARIMA modeling process for time‑series analysis, covering data loading, stationarity and white‑noise tests, model order selection, parameter estimation, diagnostic checks, and future forecasting with detailed code examples.
The article explains that attribution analysis of metric declines requires a quantitative approach, introducing Dominance Analysis—a econometric technique that decomposes regression R² into variable-specific contributions by fitting all subset models, averaging marginal effects, ranking factors, and providing a Python implementation with the dominance‑analysis package illustrated on the Boston Housing dataset.
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.
This article explores five primary correlation analysis models—rule‑based, statistical, threat‑intelligence‑based, context‑based, and big‑data‑driven—detailing their principles, typical use cases such as single‑log alerts, event‑count thresholds, multi‑value detections, temporal sequences, and how accurate log parsing underpins effective security analytics.
This article explains the fundamentals of Bayes' theorem, shows how to compute prior, likelihood, and posterior probabilities, demonstrates Bayesian A/B testing with Python code, introduces Bayesian networks for causal inference, and discusses the role of Bayesian methods in machine learning and data‑driven decision making.
