PRINCIPLES OF OR
    The following principles must be kept in mind 
     while formulating models.
     PRINCIPLE OF SIMPLICITY
    Mathematicians are of the habit of making complex 
     models whereas one should go in for simple model. It 
     means model must be kept simple models.
     UNDERSTAND THE PROBLEM, ONLY THEN APPLY THE 
     APPROPRIATE TECHNIQUE
    If we do not understand the problem properly, we cannot 
     apply the appropriate technique of operation research. 
     For example, in case of allocation of scarce resources, 
     the technique of LPP may be applied.
  
    IMPLEMENTATION
  Model must be validated before implementation 
  otherwise it can be implemented in phases for 
  validation, example Two-Phase simplex method.
  
    ANALYSIS
  A model can analyse the data but it cannot be better 
  than the information that goes into it. It means data 
  can be interpreted but cannot be generated.
  
    MODELS ARE FOR THE DECISION 
  MAKERS
  OR models are to aid the decision maker but not to 
  replace them. The decision is to be taken by the 
  management e.g.
  In LPP X1 = 0                   X2 = 9
              X1 = 6       or        X2 = 3
  Say, in the above multiple optimal solution is provided 
  by LPP giving same total profits. It is for the 
  management to decide whether to discontinue the 
  product X1 or continue to produce X1 because the 
  total profits are the profits are the same.
  
   A MODEL SHOULD NEVER BE TAKE TOO  
  LITERALLY
   It means it does not provide unique answer. In 
  sensitivity analyses, or under simulation, we can find 
  still better results.
  
   FLEXIBLE MODEL
  A model should be flexible enough to incorporate 
  changes. It should give range where one solution is 
  valid as in case of sensitivity analysis
  
    SEEK COOPERATION FROM OR EXPERTS
  As we know the tested results of the model are 
  implemented. It is to be done by OR expert results. 
  
   USE OF COMPUTERS
  Use of computers is made where ever possible in case 
  of implementation of models. OR techniques are 
  usually complex and only computers can solve them. 
  Hence, steps should be clearly stated to unable the OR 
  researcher to develop computer software for future 
  implementation of the techniques.
                                   
                  THANK 
      YOU