, Indoor air quality data: comparison to the global model & parameter selection Interpretation Comparison to the global model: (small) RMSE improvement in comparison with
, Selection of the optimal number of model dimensions, p.1
, Indoor air quality data: comparison to the global model & parameter selection Interpretation Comparison to the global model: (small) RMSE improvement in comparison with
, Selection of the optimal number of model dimensions: 1 Aim: improve the prediction, Meaningful criterion to minimize, Parameters (number of dimensions and clusters) obtained through cross-validation
, Useful tools to deal with real data especially in biology (e.g., different risk factors of a disease according to sub-populations)
Programs will be transformed into a R package. Aim: improve the prediction, Meaningful criterion to minimize, Parameters (number of dimensions and clusters) obtained through cross-validation ,
, Useful tools to deal with real data especially in biology (e.g., different risk factors of a disease according to sub-populations)
, Perspectives Any other supervised multiblock methods can be included in the algorithm, Next step: allow specific clusters (and dimensions) for each block
, Gilbert Saporta (1) & Hervé Abdi
,
, French Agency for food, occupational and health safety (Anses)
, Correspondence Analysis and Related Methods, pp.20-23