Related ideas

• Naive Bayes
• Kalman filter
• Make available the permuted clustering table used to construct heatmaps.
• Tests on contingency tables: add p-values in significance by cells
• integrate Correspondence Analysis in the MFA (Multiple Factor Analysis) to treat Frequency tables
• Combination of Factorial Methods and Cluster Analysis
• Empirical Bayes - moderated t-tests in differential expression
• Penalized regression (logistic AND linear)
• improve product characterisation