statistics

统计分析方法、假设检验和概率论在数据分析中的应用

📖 中文导读:统计分析方法、假设检验和概率论在数据分析中的应用 以下正文为该技能的原始 SKILL.md 内容(英文),上方安装命令可直接使用。

name: statistics description: Statistical analysis methods, hypothesis testing, and probability for data analytics version: "2.0.0" sasmp_version: "2.0.0" bonded_agent: 03-statistical-analysis-expert bond_type: PRIMARY_BOND

Skill Configuration

config: atomic: true retry_enabled: true max_retries: 3 backoff_strategy: exponential numerical_precision: high

Parameter Validation

parameters: skill_level: type: string required: true enum: [beginner, intermediate, advanced] default: beginner focus_area: type: string required: false enum: [descriptive, inferential, probability, regression, experiments, all] default: all tool_preference: type: string required: false enum: [python, r, excel, all] default: python

Observability

observability: logging_level: info metrics: [calculation_accuracy, test_validity, model_fit]

Statistics Skill

Overview

Master statistical concepts and methods essential for data analysis, from descriptive statistics to advanced inferential techniques.

Core Topics

Descriptive Statistics

  • Measures of central tendency (mean, median, mode)
  • Measures of dispersion (variance, standard deviation, IQR)
  • Data distributions and skewness
  • Percentiles and quartiles

Inferential Statistics

  • Sampling methods and sample size determination
  • Confidence intervals
  • Hypothesis testing (t-tests, chi-square, ANOVA)
  • P-values and statistical significance

Probability

  • Basic probability rules
  • Probability distributions (normal, binomial, Poisson)
  • Bayes' theorem
  • Expected value and variance

Regression Analysis

  • Linear regression
  • Multiple regression
  • Logistic regression
  • Model validation and diagnostics

Learning Objectives

  • Apply descriptive statistics to summarize data
  • Conduct hypothesis tests for business decisions
  • Build and interpret regression models
  • Communicate statistical findings effectively

Error Handling

Error Type Cause Recovery
Sample too small Insufficient data Increase sample or use bootstrap
Assumption violated Data doesn't fit test Use non-parametric alternative
Multicollinearity Correlated predictors Remove or combine variables
Outliers Extreme values Investigate or use robust methods
P-hacking Multiple testing Apply Bonferroni correction
  • programming (for implementing statistical models)
  • visualization (for presenting statistical insights)
  • advanced (for machine learning)