Developing Two Combined Methods Based on Discriminant Analysis to Monitor Distribution of narcotic drugs in Iran

Nowadays, narcotic drugs problem is one of the most important social damages in Iran. Addiction can lead to some destructive effects in the society including economic, cultural, mental, moral and legal effects. Hence, monitoring the smuggling of narcotic drugs in Iran is inevitable and improves the community health. Distribution of the narcotic drugs can be characterized by two correlated quality characteristics including the weights of discovered narcotic drugs and the number of arrested smugglers by the police. It is assumed that the first quality characteristic is a variable and follows a normal distribution whereas the second is an attribute and follows a Poisson distribution. The positive correlation also exists between the variable and attribute quality characteristics. In this paper, two methods based on the discriminant analysis are proposed for simultaneous monitoring of correlated variable and attribute quality characteristics. The first method is based on the combination of discriminant analysis technique and artificial neural network and the second is based on the discriminant analysis technique and exponential weighted moving average (EWMA) control chart. The performance of the proposed methods in detecting different step shifts in the quality characteristics is evaluated through simulation in terms of average run length criterion. Finally, a numerical example is presented to elaborate the proposed methods. The results confirm the high performance of both methods for monitoring purpose and can be implemented by the police to monitor the distribution of the narcotic drugs in the country.