Demand Modeling
We introduce the binomial distribution, which can be used to model uncertain customer demand when the number of customers and their likelihood of making a purchase are known. In many problem settings, the store does not know the exact number of customers. In such cases, the Poisson distribution may be more appropriate for modelling customer demand. In particular, the Poisson distribution makes use of information regarding average sales to model customer demand. If demand variance is also be known, the normal distribution may be used.
Binomial distribution (16 questions)
Suppose there are 3 customers A, B and C. Each customer will purchase your product with probability 0.1. Observe that there are 8 possible scenarios:
Poisson distribution (6 questions)
Suppose that a store sells an average of 10 products each day. If we were to model demand using the binomial distribution where n = 100 (i.e., 100 potential customers), we should set the probability of a customer purchasing a product to be 0.1 in order to be consistent with the fact that the store sells an average of 10 products daily.
Normal distribution (19 questions)
The normal distribution is frequently used to model demand in practice. One reason for this is that the normal distribution is defined by two parameters (i.e., mean and variance) which are easily estimated from historical demand data.