Fuzzy sets and fuzzy logic can be used for efficient data classification by fuzzy rules and fuzzy classifiers. This paper presents an application of genetic programming to the evolution of fuzzy classifiers based on extended Boolean queries. Extended Boolean queries are well known concept in the area of fuzzy information retrieval. An extended Boolean query represents a complex soft search expression that defines a fuzzy set on the collection of searched documents. We interpret the data mining task as a fuzzy information retrieval problem and we apply a proven method for query induction from data to find useful fuzzy classifiers. The ability of the genetic programming to evolve useful fuzzy classifiers is demonstrated on two use cases in which we detect faulty products in a product processing plant and discover intrusions in a computer network.

Systemic lupus erythematosus (SLE) is a remarkably heterogeneous autoimmune disease. Despite tremendous efforts, our knowledge of serum protein patterns in severe SLE phenotypes is still limited. We investigated the serum protein pattern of SLE, with special emphasis on irreversible organ damage and active lupus nephritis (LN) as assessed by renal Systemic Lupus Erythematosus Disease Activity Index. We used proximity extension immunoassay (PEA, Proseek Multiplex, Olink) to assess the serum levels of ninety-two inflammation-related proteins in Czech patients with SLE (n = 75) and age-matched healthy control subjects (n = 23). Subgroup analysis was carried out on the basis of organ damage (with/without, 42/33) and biopsy-proven LN (with/without, 27/48; active LN, n = 13; inactive LN, n = 14). Of thirty deregulated proteins between SLE and the healthy controls (P corr < 0.05), the top upregulated proteins in SLE were sirtuin 2, interleukin 18 (IL18), and caspase 8 (P corr < 0.0006). Of these, sirtuin 2 and caspase 8 had not yet been reported with SLE. Elevated levels of IL8, CCL2/MCP1, CCL11, and MMP10 (P corr < 0.05) were detected in patients with organ damage for which the serum levels of CCL11 and MMP10 were particularly informative in organ damage prediction. Comparing patients based on LN, elevated levels of CSF1, sIL15RA, sCD40, sCX3CL1, caspase 8, sIL18R1, bNGF, and GDNF (P corr < 0.05) were detected in active LN. Except GDNF, all LN-associated markers showed usefulness in prediction of active renal disease. This highly sensitive PEA analysis identified the serum pattern of SLE, organ damage, and active LN, with many novel candidate proteins detected. Their exact role and suitability as biomarkers in SLE deserve further investigation.