SNAD transient miner: Finding missed transient events in ZTF DR4 using k-D trees

We report the automatic detection of 11 transients (7 possible supernovae and 4 active galactic nuclei candidates) within Zwicky Transient Facility fourth data release (ZTF DR4), all them observed in 2018 absent from public catalogs. Among these, three were not part ZTF alert stream. Our transient mining strategy employs 41 physically motivated features extracted both real light curves four simulated curve models (SN Ia, SN II, TDE, SLSN-I). These are input to a k-D tree algorithm, which we calculate 15 nearest neighbors. After pre-processing selection cuts, our dataset contained approximately million objects among visually inspected 105 closest neighbors seven brightest, most well-sampled simulations, comprising 89 unique DR4 sources. result illustrates potential coherently incorporating domain knowledge learning algorithms, is one guiding principles directing SNAD team. It also demonstrates that DR suitable testing ground for algorithms aiming prepare next generation astronomical data.