the present invention relates to improved methods of identifying the persons daily activities. includes a new method for identifying activities that use context to determine the current activity, and improves the accuracy of the identification. the method uses the sampled data linear and gravity acceleration levels. these data provide the motion sensor, which is directly or indirectly supported by the subject.processor for the extraction of context information is calculated in several contexts. processor for modelling context, with the assistance of machine learning to model for each value of the context. the activity of the entity are identified by combining, viz., the results of several models of context. the invention allows the application to identify activities that determine the activity of the entity.the application may include the information mechanism of the alarm when detecting the fall. you can also include, inter alia: application for general application to daily activity recognition, applications for aid in ambient assisted living, etc. the technical u\u010dinekizuma allows control of the smart home and the more accurate detection of the fall, which triggers an automatic emergency call.Pričujoči izum je povezan z izboljšanimi metodami prepoznavanja dnevnih aktivnosti osebe. Vključuje novo metodo za prepoznavanje aktivnosti, ki uporablja kontekst trenutne aktivnosti, da jo določi in izboljša točnost njenega prepoznavanja. Metoda uporablja vzorčene podatke o linearnih in gravitacijskih pospeških. Te podatke poda senzor gibanja, katerega posredno ali neposredno podpira subjekt. Procesor za ekstrakcijo konteksta izračuna več kontekstov iz podatkov. Procesor za modeliranje konteksta se spomočjo strojnega učenja nauči modela za vsako vrednost konteksta. Aktivnost subjekta se prepozna z združevanjem, t. j. združenje rezultatov več modelov konteksta. Izum omogoča aplikacijo za prepoznavanje aktivnosti, ki določi aktivnost subjekta. Aplikacija lahko vključuje me
