x********u
发帖数: 64 | 1 there are 4 products A/B/C/D in a direct mail marketing campaign
proespects can enroll in one of the product or 2, or 3, or all 4 products or
not to enroll in any of the products
what kind of model to be used to predict enrollment of the products?
what would be the model outcome?
Thanks | h***x
发帖数: 586 | 2 具体啥模型，根据你想干什么
比方你想预测哪些customer是会买1个产品，2个产品，3个产品和4个产品，就是
ordinal logistic regression了。
出来4个equations，变量都一样，intercept不同。分别预测的是买4个产品的概率，3
个以上产品的概率，2个以上产品的概率和买产品的概率
如果牵涉到购买具体产品的模型，可能会是binary/multinomial logistical
regression
target的levels如果太多，modeling不好控制，需要注意
or
【在 x********u 的大作中提到】
: there are 4 products A/B/C/D in a direct mail marketing campaign
: proespects can enroll in one of the product or 2, or 3, or all 4 products or
: not to enroll in any of the products
: what kind of model to be used to predict enrollment of the products?
: what would be the model outcome?
: Thanks
|
|
