Table 2 Main results on the real-world dataset, SACHS. Unconstrained versions of causal discovery learning yield relatively low performance, while the NoTears algorithms are the two best. MetaNoTears-L1 is the best-performing algorithm and shows improvement in the metrics. Abbreviations: FDR (false discovery rate), TPR (true positive rate), FPR (false positive rate), SHD (structural Hamming distance), and NNZ (number of non-zero entries)

From: Developing a novel causal inference algorithm for personalized biomedical causal graph learning using meta machine learning