| Popis: |
Strojno učenje je učenje računala prepoznavanju uzoraka koji se ponavljaju u setu za treniranje kojeg smo mu poslali, da bi prema njima mogao predvidjeti izlazne rezultate, tj. neko buduće ponašanje.. Model se uči na danom setu za treniranje, nakon kojih može uspješno procesuirati i nove, nepoznate podatke te predvidjeti stanja. U strojnom učenju postoji velik broj tipova modela, od kojih smo odabrali linearnu regresiju. Najprije podatke razvrstavamo na set za treniranje i testiranje, zatim popunjavamo praznine u podatcima interpolacijom, skaliramo varijable, obavljamo linearnu regresiju, inverzno skaliramo te na kraju računamo srednju kvadratnu grešku i grafički prikazujemo izlaznu varijablu. Machine learning is defined as learning a computer to recognize repetitive patterns in a training set that we send it, so that it can predict the output, ie some future behavior. The model is trained on a given training set, after which it can successfully process new, unknown data and predict conditions. In machine learning, there is a large number of types of models, from which we have chosen linear regression. We first sort the data into a training and testing set, then fill in the gaps in the data by interpolation, scale the variables, perform linear regression, inversely scale, and finally calculate the mean square error and graphically display the output variable. |
