Heisenberg uncertainly principle can be discussed by using the pattern classification technique of Artificial Neutral Networks with Back propagation learning rue. In this process a suitable arrangement can be constructed that contains the control of two experiments, one of which is designed to measure the position and other one is designed to measure the momentum of a sub-atomic particle (the electron). The Control system will determine either the position of momentum of electron, which less uncertain for any wavelength of light at any instant and start measuring the less uncertain quantity with the corresponding experiment.
The measurement of spin-1/2 observables using a Stern-Gerlach type device is studied. A magnetic field with a dominant dipole component leads to a measurement of spin in one direction. The measurement is shown to be feasible (in principle) for electrons as well as for neutral particles. A quadrupole field leads to a joint measurement of two incompatible spin observables. Again, both the electron and the neutral case are presented. The accuracy of the joint measurement is compared to the limit imposed by the uncertainty principle.