This thesis is concerned with the application of artificial neural network(ANN) methodology to the operation of multivariate control charts when there are two or more correlated quality variables. Hotelling's $T^2$ control chart is used to monitor overall process changes and the ANN method is used as a supplementary analysis to pin-point out-of-control variables when out-of-control signals appear in the $T^2$ chart. The method of using the ANN is compared with the Bonferroni type simultaneous confidence intervals in terms of the rate of correct identification of out-of-control variables when the number of variables is 2 or 4. The results show that the proposed method performs better than the Bonferroni method when the quality variables are highly correlated and the two methods are comparable when the correlations are relatively small.