In this thesis nitrite and bromide ions were determined simultaneously based on the catalytic effects of these anions on pyrogallal red reaction, using principle component analysis and artificial neural networks. In this work, principle component analysis was used for dimension reduction of the data. The values of mean relative error for bromide ion in calibration set was 8.10% and for nitrite ion 7.3%. Therefore this work propose a method for determining bromide and nitrite ions in concentration range of 0.50 – 110 microgram per milliliter and 20-2000 nanogram per milliliter, respectively. In other part of this thesis, piroxicam and pyridoxine were determined using spectrofluorimetry based on the parallel factor analysis (PARAFAC) method. In this way, fluorescence emission spectrums were obtained from 330-540 nm, for excitation wavelengths from 280-320 nm. Consequently, one matrix was obtained for each sample, and three dimensional matrix was obtained for all of the samples. Mean relative errors for real samples were 3.33% for piroxicam and 6.50% for pyridoxine. As a result this analysis method can be used to determine two mentioned components in concentration range of 0.0025-10.0 microgram per milliliter and 0.001-0.50 microgram per milliliter, for piroxicam and pyridoxine respectively.