Use Of Neural Networks For The Identification Of New Z >= 3.6 Qsos From First-Sdss Dr5

We aim to obtain a complete sample of z ≥ 3. 6 radio QSOs from FIRST sources having star-like counterparts in the SDSS DR5 photometric survey (r AB ≤ 20. 2). The starting sample of FIRST–DR5 pairs includes 4,250 objects with DR5 spectra, 52 of these being z ≥ 3. 6 QSOs. Simple supervised neural networks, trained on these sources, using optical photometry and radio data, are very effective for identifying high-z QSOs, yielding 96% completeness and 62% efficiency. Applying these networks to the 4,415 FIRST–DR5 sources without DR5 spectra we found 58 z ≥ 3. 6 QSO candidates. We obtained spectra of 27 of them, confirming 17 as high-z QSOs. Spectra of 13 additional candidates from the literature and SDSS DR6 revealed seven more z ≥ 3. 6 QSOs, giving an overall efficiency of 60% (24/40). None of the non-candidates with spectra from NED or DR6 is a z ≥ 3. 6 QSO, consistently with a high completeness. The initial sample of high-z QSOs is increased from 52 to 76 sources (a factor 1.46). From the new identifications and candidates we estimate an incompleteness of SDSS for the spectroscopic classification of FIRST 3. 6 ≤ z ≤ 4. 6 QSOs of 15% for r ≤ 20. 2.