說是看輸出,不看輸入。參數見內

回答: 他家主要看什麽?randomness2023-09-06 19:47:28

33%: difference between observed graduate median earnings and a statistical model's predicted median earnings

17%: years to pay off the average net price, which uses the average net cost of attendance and the estimated value added at the median earnings over a HS graduate's median earnings in that college's state to calculate how long it took for the investment to break even

20%: graduation rate in comparison to another statistical model, blended with absolute graduation rate

5%: quality and frequency of learning opportunities as per a survey

5%: quality and frequency of career preparation opportunities as per a survey

5%: satisfaction with learning facilities as per a survey

5%: would-you-recommend/satisfaction scores as per a survey

5%: opportunities to interact with people from different backgrounds as per a survey

1.5%: the probability that two students or faculty members would be of a different ethnicity

1.5%: proportion of Pell Grants

1%: proportion of disabled students

1%: proportion of students from outside the US

所有跟帖: 

好像和紫檀理念比較接近。看畢業以後收入和學費投資回報率 -randomness- 給 randomness 發送悄悄話 (0 bytes) () 09/06/2023 postreply 19:58:59

性價比 -yzhl888- 給 yzhl888 發送悄悄話 (0 bytes) () 09/06/2023 postreply 20:08:34

看加成,對學生的提升度 -mjnew- 給 mjnew 發送悄悄話 (0 bytes) () 09/06/2023 postreply 20:12:45

問題是數據的可靠性太差 -CatcherInTheRye- 給 CatcherInTheRye 發送悄悄話 (97 bytes) () 09/06/2023 postreply 20:21:01

看性價比是對的,IIT還可以吧,大部分都是理工科,學費也不是很高 -Amita- 給 Amita 發送悄悄話 (0 bytes) () 09/06/2023 postreply 20:29:38

結果前麵反正還是那幾家 -randomness- 給 randomness 發送悄悄話 (0 bytes) () 09/06/2023 postreply 20:34:53

用了太多的model, 就用median earning/graduation rate真實數據不好嗎?而且那個33%為 -Lionkingmom- 給 Lionkingmom 發送悄悄話 (0 bytes) () 09/07/2023 postreply 05:31:57

什麽用真實的收入和模型預測收入差別來計算?這個差別能說明什麽? 不理解 -Lionkingmom- 給 Lionkingmom 發送悄悄話 (0 bytes) () 09/07/2023 postreply 05:35:14

請您先登陸,再發跟帖!