Statistics

Content

• Summary of Inferential Statistics : Summarize different statistical methods according to continuous / category data, two or more samples, independent / dependent samples, etc.
• Correlation : 多筆資料的連續性相關分析，並依資料是否為常態假設或符合中央極限定理而有不同。
  • Pearson correlation : based on Gussian distribution or central limit theorem calculate how correlated between two matrixes.
  • Spearman Rank Correlation : Spearman calculates correlation between two not in Gussian distribution or central limit theorem.
• Normalization : 正規化或移除機器背景等問題，促使資料分布相似的方法，又可以依據是否有母體參考而分成母數(參數)與無母數(非參數)的正規化處理方法。
  • Quantile Normalization : It could be used to scale the data into similiar range within the group.
• Fisher's exact test : A rigorous method replaces chi-squared test in large samples.
• Hypergeometric test : A rigorous method uses in analyzing high-dimensional tree-level data.

Resource