航空公司的售票营销策略.docVIP

航空公司的售票营销策略 航空公司的售票营销策略 PAGE 航空公司的售票营销策略 应用统计学案例3，学时：36小时， 案例规模：中 类型：选编 来源：Statistics For Business And Economics,David R. Anderson South-Western Thomson Learning,1998 Estimating the chances of “No shows” and Is Overbooking Worth it? Transamerican Airlines has opened a new daily flight from Chicago to Boise, Idaho. It is so popular that its 75 seats have all been reserved in its first 20 flights Unfortunately, in each flight some of the passengers failed to show up. So that the plane left with some empty seats and lost revenue. The number of “no shows” varied from 2 on the best flight to 11 on the worst flight. Here are the details: Number of “No Shows” x Frequency 2 3 4 5 6 7 8 9 10 11 1 4 0 4 2 5 1 1 0 2 20 Flights To reduce the number of empty seats, the airline can sell more than 75 reservations, of course. But that introduces a new risk of overbooking, if more than 75 people show up with reservations. In order to properly balance its risks, the airline needs to estimate the chances of various no shows”. We therefore imagine collecting data for many flights (under similar conditions) so that the relative frequencies would settle down to probabilities p(x). Let us estimate one of these probabilities-for example, the chance of exactly 3 “no shows”, p(3). To estimate this long-run relative frequency p(3). The shortrun relative frequency in the table above is 4/20=. Do you think this estimate is too high, or too low Why Graph the relative frequency distribution of X from the table above. Sketch a smooth curve through it-a crude probability of p(3) For a more refined model, let p denote the proportion of passenger who are “no shows”, Then assume that a flight consists of n=75 randomly chosen passengers. So that X has binomial distribution. To estimate p. how many passengers were there altogether in the 20flights And how many “no shows” So what is your estimate of p Now estimate p(3) using the binomial distribution. Graph it alongsi