在用frontier4.1中的BC1995模型（模型2）计算面板数据时，结果如何解释？


Output from the program FRONTIER (Version 4.1c)

instruction file = kcsjm2.ins  
data file =        kcsjm2.dta  

Tech. Eff. Effects Frontier (see B&C 1993)
The model is a production function
The dependent variable is not logged

the ols estimates are :
                 coefficient     standard-error    t-ratio
  beta 0         0.22591901E+05  0.35217695E+05  0.64149290E+00
  beta 1         0.11619231E+01  0.10569134E+01  0.10993551E+01
  beta 2         0.11558298E+01  0.17379257E-01  0.66506281E+02
  beta 3         0.66236876E-01  0.83244913E-01  0.79568677E+00
  sigma-squared  0.73858757E+11
log likelihood function =  -0.25892648E+04
the estimates after the grid search were :
  beta 0         0.28286612E+06
  beta 1         0.11619231E+01
  beta 2         0.11558298E+01
  beta 3         0.66236876E-01
  delta 1        0.00000000E+00
  sigma-squared  0.14001307E+12
  gamma          0.76000000E+00


iteration =     0  func evals =     20  llf = -0.25768253E+04
     0.28286612E+06 0.11619231E+01 0.11558298E+01 0.66236876E-01 0.00000000E+00
     0.14001307E+12 0.76000000E+00
gradient step
iteration =     5  func evals =     44  llf = -0.25652833E+04
     0.28286612E+06-0.58721959E+00 0.12076266E+01 0.44433034E-01 0.11275841E+01
     0.14001307E+12 0.87829753E+00
iteration =    10  func evals =    142  llf = -0.25633269E+04
     0.28286612E+06-0.19481303E+01 0.12412228E+01 0.97222537E-01 0.21621906E+01
     0.14001307E+12 0.91439169E+00
search failed. fn val indep of search direction
iteration =    11  func evals =    144  llf = -0.25633269E+04
     0.28286612E+06-0.19481303E+01 0.12412228E+01 0.97222537E-01 0.21621906E+01
     0.14001307E+12 0.91439169E+00

the final mle estimates are :
                 coefficient     standard-error    t-ratio
  beta 0         0.28286612E+06  0.10002412E+01  0.28279791E+06
  beta 1        -0.19481303E+01  0.96951344E+00 -0.20093897E+01
  beta 2         0.12412228E+01  0.20635989E-01  0.60148451E+02
  beta 3         0.97222537E-01  0.79695107E-01  0.12199311E+01
  delta 1        0.21621906E+01  0.56144440E+00  0.38511214E+01
  sigma-squared  0.14001307E+12  0.10000000E+01  0.14001307E+12
  gamma          0.91439169E+00  0.24391378E-01  0.37488316E+02
log likelihood function =  -0.25633269E+04
LR test of the one-sided error =   0.51875829E+02
with number of restrictions = 2
[note that this statistic has a mixed chi-square distribution]
number of iterations =     11
(maximum number of iterations set at :   100)
number of cross-sections =     31
number of time periods =      6
total number of observations =    186
thus there are:      0  obsns not in the panel

covariance matrix :
  0.10004824E+01  0.12047308E-01 -0.90968524E-04 -0.52830483E-03 -0.29906358E-02
-0.34834248E-15  0.26294225E-03
  0.12047308E-01  0.93995632E+00 -0.13992620E-01 -0.36154807E-01 -0.33354104E+00
  0.78740787E-11 -0.96356508E-02
-0.90968524E-04 -0.13992620E-01  0.42584405E-03 -0.27071362E-03  0.78051863E-02
-0.46455443E-13  0.27786369E-03
-0.52830483E-03 -0.36154807E-01 -0.27071362E-03  0.63513100E-02  0.14603726E-01
  0.49512128E-12  0.16656540E-03
-0.29906358E-02 -0.33354104E+00  0.78051863E-02  0.14603726E-01  0.31521981E+00
  0.51806195E-12  0.68671596E-02
-0.34834248E-15  0.78740787E-11 -0.46455443E-13  0.49512128E-12  0.51806195E-12
  0.10000000E+01  0.73861353E-12
  0.26294225E-03 -0.96356508E-02  0.27786369E-03  0.16656540E-03  0.68671596E-02
  0.73861353E-12  0.59493932E-03