比较线性模型和Probit模型、Logit模型文档格式.doc
《比较线性模型和Probit模型、Logit模型文档格式.doc》由会员分享,可在线阅读,更多相关《比较线性模型和Probit模型、Logit模型文档格式.doc(5页珍藏版)》请在冰豆网上搜索。
330
71
267
6
379
39
328
72
266
7
378
40
73
263
8
41
74
261
9
376
42
321
75
260
10
371
43
76
256
11
362
44
318
77
252
12
45
78
13
361
46
316
79
245
14
359
47
308
80
243
15
358
48
81
242
16
356
49
304
82
241
17
50
303
83
239
18
355
51
84
235
19
354
52
299
85
232
20
53
297
86
228
21
353
54
294
87
219
22
350
55
293
88
23
349
56
89
214
24
57
292
90
210
25
348
58
291
91
204
26
347
59
92
198
27
60
287
93
189
28
344
61
286
94
188
29
339
62
95
182
30
338
63
282
96
166
31
64
97
123
32
336
65
33
334
66
278
定义变量SCORE:
考生考试分数;
Y:
考生录取为1,未录取为0。
上图为样本观测值。
1.线性概率模型
根据上面资料建立模型
用Eviews得到回归结果如图:
DependentVariable:
Y
Method:
LeastSquares
Date:
12/10/10Time:
20:
Sample:
197
Includedobservations:
97
Variable
Coefficient
Std.Error
t-Statistic
Prob.
C
-0.847407
0.159663
-5.307476
0.0000
0.003297
0.000521
6.325970
R-squared
0.296390
Meandependentvar
0.144330
AdjustedR-squared
0.288983
S.D.dependentvar
0.353250
S.E.ofregression
0.297866
Akaikeinfocriterion
0.436060
Sumsquaredresid
8.428818
Schwarzcriterion
0.489147
Loglikelihood
-19.14890
F-statistic
40.01790
Durbin-Watsonstat
0.359992
Prob(F-statistic)
0.000000
参数估计结果为:
-0.847407+0.003297
Se=(0.159663)(0.000521)
t=(-5.307476)(6.325970)
p=(0.0000)(0.0000)
预测正确率:
Forecast:
YF
Actual:
Forecastsample:
RootMeanSquaredError
0.294780
MeanAbsoluteError
0.233437
MeanAbsolutePercentageError
8.689503
TheilInequalityCoefficient
0.475786
BiasProportion
VarianceProportion
0.294987
CovarianceProportion
0.705013
2.Logit模型
ML-BinaryLogit(Quadratichillclimbing)
21:
Convergenceachievedafter11iterations
Covariancematrixcomputedusingsecondderivatives
z-Statistic
-243.7362
125.5564
-1.941248
0.0522
0.679441
0.350492
1.938536
0.0526
0.115440
0.123553
1.266017
0.176640
-3.992330
Hannan-Quinncriter.
0.145019
Restr.loglikelihood
-40.03639
Avg.loglikelihood
-0.041158
LRstatistic(1df)
72.08812
McFaddenR-squared
0.900282
Probability(LRstat)
ObswithDep=0
Totalobs
ObswithDep=1
得Logit模型估计结果如下
pi=F(yi)=拐点坐标(358.7,0.5)
其中Y=-243.7362+0.6794X
预测正确率
0.114244
0.025502
1.275122
0.153748
0.025338
0.974662
3.Probit模型
ML-BinaryProbit(Quadratichillclimbing)
-144.4560
70.19809
-2.057833
0.0396
0.402868
0.196186
2.053504
0.0400
0.116277
0.122406
1.284441
0.175493
-3.936702
0.143872
Avg.lo