This paper presents a novel method for the selection of appropriate location and calculation of the size of distributed generation (DG) and sizing of shunt capacitors in radial distribution systems. This methodology is based on wavelet neural network (WNN) which requires less computational burden compared to traditional ones. The proposed optimization technique takes into consideration minimizing the cost of peak power, reducing energy loss and improving voltage stability index. The proposed method of optimization is then compared with the golden section search (GSS), grid search (GS), and Acharya's heuristic method (AHM) on different IEEE radial buses, i.e. 12 bus, 12 bus with DG, 33 bus, 33 bus with DG. The results demonstrates that WNN provides more economical solution by reducing power losses, energy loss, total requiring capacitive compensation and presents a good improvement in nodes voltage.