анализ данных и построение модели, позволяющей оценивать стоимость квартиры на вторичном рынке жилья
Заказать уникальную курсовую работу- 48 48 страниц
- 3 + 3 источника
- Добавлена 22.01.2017
- Содержание
- Часть работы
- Список литературы
- Вопросы/Ответы
I. Формулировка задачи 3
2. Анализ данных и выбор модели 4
2.1 Линейная модель уравнения регрессии 5
2.2 Линейная в логарифмах модель уравнения регрессии 9
2.3 Полулогарифмическая модель уравнения регрессии 16
3. Оценка значимости полученных уравнений 20
3.1 Оценка зависимости цены квартиры от различных факторов 20
3.2 Тестирование моделей 22
Выводы 27
Литература 28
Приложение 29
Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- ln_in1 | .5535358 .0555342 9.97 0.000 .4428563 .6642152 ln_in5 | .4166709 .0647407 6.44 0.000 .2876428 .5456991 _cons | 1.71318 .14108 12.14 0.000 1.432008 1.994352------------------------------------------------------------------------------. hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 4.00 Prob > chi2 = 0.0454. regress lg_cost ln_in3 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 1, 74) = 319.49 Model | 11.9217478 1 11.9217478 Prob > F = 0.0000 Residual | 2.76128207 74 .037314623 R-squared = 0.8119-------------+------------------------------ Adj R-squared = 0.8094 Total | 14.6830298 75 .195773731 Root MSE = .19317------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- ln_in3 | .9476271 .053016 17.87 0.000 .8419904 1.053264 _cons | -.835674 .2225597 -3.75 0.000 -1.279134 -.3922142------------------------------------------------------------------------------. hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 1.29 Prob > chi2 = 0.2564. regress lg_cost in1 in2 in4 in5 in6 in7 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 6, 69) = 48.22 Model | 11.8556239 6 1.97593732 Prob > F = 0.0000 Residual | 2.82740592 69 .040976897 R-squared = 0.8074-------------+------------------------------ Adj R-squared = 0.7907 Total | 14.6830298 75 .195773731 Root MSE = .20243------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- in1 | .0974487 .0421487 2.31 0.024 .0133643 .181533 in2 | .001245 .0506131 0.02 0.980 -.0997254 .1022154 in4 | .0112271 .0022916 4.90 0.000 .0066554 .0157987 in5 | .0183765 .0048243 3.81 0.000 .0087522 .0280008 in6 | .0140859 .0518186 0.27 0.787 -.0892894 .1174612 in7 | -.0078666 .0042375 -1.86 0.068 -.0163201 .000587 _cons | 2.238753 .089357 25.05 0.000 2.060491 2.417016------------------------------------------------------------------------------. hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 2.51 Prob > chi2 = 0.1130. regress lg_cost ln_in1 in2 ln_in3 ln_in5 in6 ln_in7 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 6, 69) = 52.69 Model | 12.0525157 6 2.00875262 Prob > F = 0.0000 Residual | 2.63051411 69 .038123393 R-squared = 0.8208-------------+------------------------------ Adj R-squared = 0.8053 Total | 14.6830298 75 .195773731 Root MSE = .19525------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- ln_in1 | -.0329018 .1154151 -0.29 0.776 -.2631486 .1973451 in2 | -.0016812 .0499283 -0.03 0.973 -.1012856 .0979231 ln_in3 | .998441 .1805438 5.53 0.000 .6382658 1.358616 ln_in5 | -.0054268 .0969092 -0.06 0.956 -.1987555 .1879018 in6 | .0479156 .0499756 0.96 0.341 -.0517829 .1476142 ln_in7 | -.0615344 .0374959 -1.64 0.105 -.1363367 .0132679 _cons | -.8918956 .5184027 -1.72 0.090 -1.926081 .1422896------------------------------------------------------------------------------. vif Variable | VIF 1/VIF -------------+---------------------- ln_in3 | 11.35 0.088097 ln_in1 | 6.78 0.147486 ln_in5 | 3.52 0.284302 in6 | 1.17 0.851125 in2 | 1.17 0.852737 ln_in7 | 1.06 0.942794-------------+---------------------- Mean VIF | 4.18. hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 1.95 Prob > chi2 = 0.1628. stepwise, pr (0.05) : regress lg_cost in1 in2 in3 in5 in6 in7 begin with full modelp = 0.9748 >= 0.0500 removing in2p = 0.8555 >= 0.0500 removing in6p = 0.2492 >= 0.0500 removing in5p = 0.0865 >= 0.0500 removing in7 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 2, 73) = 149.72 Model | 11.805049 2 5.90252452 Prob > F = 0.0000 Residual | 2.87798082 73 .039424395 R-squared = 0.8040-------------+------------------------------ Adj R-squared = 0.7986 Total | 14.6830298 75 .195773731 Root MSE = .19856------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- in1 | .0773559 .034649 2.23 0.029 .0083006 .1464113 in3 | .0107338 .0012631 8.50 0.000 .0082164 .0132512 _cons | 2.155491 .0616769 34.95 0.000 2.032569 2.278413------------------------------------------------------------------------------. hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 0.11 Prob > chi2 = 0.7388. regress lg_cost in1 in4 in5 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 3, 72) = 94.69 Model | 11.7140998 3 3.90469994 Prob > F = 0.0000 Residual | 2.96893003 72 .041235139 R-squared = 0.7978-------------+------------------------------ Adj R-squared = 0.7894 Total | 14.6830298 75 .195773731 Root MSE = .20306------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- in1 | .0983293 .0415578 2.37 0.021 .0154854 .1811732 in4 | .0115103 .0022651 5.08 0.000 .0069948 .0160257 in5 | .0163053 .0046205 3.53 0.001 .0070944 .0255162 _cons | 2.169695 .0682718 31.78 0.000 2.033597 2.305792------------------------------------------------------------------------------. hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 1.33 Prob > chi2 = 0.2485. boxcox cost ln_ in3, model (theta)ln_ ambiguous abbreviationr(111);. regress lg_cost ln_in3 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 1, 74) = 319.49 Model | 11.9217478 1 11.9217478 Prob > F = 0.0000 Residual | 2.76128207 74 .037314623 R-squared = 0.8119-------------+------------------------------ Adj R-squared = 0.8094 Total | 14.6830298 75 .195773731 Root MSE = .19317------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- ln_in3 | .9476271 .053016 17.87 0.000 .8419904 1.053264 _cons | -.835674 .2225597 -3.75 0.000 -1.279134 -.3922142------------------------------------------------------------------------------. boxcox lg_cost ln_ in3, model (theta)ln_ ambiguous abbreviationr(111);. boxcox lg_cost ln_in3, model (theta)Fitting comparison modelIteration 0: log likelihood = -45.365773 Iteration 1: log likelihood = -44.605227 Iteration 2: log likelihood = -44.60491 Iteration 3: log likelihood = -44.60491 Fitting full modelIteration 0: log likelihood = 18.132125 Iteration 1: log likelihood = 20.424381 Iteration 2: log likelihood = 20.429501 Iteration 3: log likelihood = 20.429501 Number of obs = 76 LR chi2(2) = 130.07Log likelihood = 20.429501 Prob > chi2 = 0.000------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. z P>|z| [95% Conf. Interval]-------------+---------------------------------------------------------------- /lambda | .5996746 1.405472 0.43 0.670 -2.154999 3.354349 /theta | -.3065964 .8161574 -0.38 0.707 -1.906235 1.293043------------------------------------------------------------------------------Estimates of scale-variant parameters---------------------------- | Coef.-------------+--------------Notrans | _cons | .090743-------------+--------------Trans | ln_in3 | .381472-------------+-------------- /sigma | .0423141------------------------------------------------------------------------------------------- Test Restricted H0: log likelihood chi2 Prob > chi2---------------------------------------------------------------theta=lambda = -1 19.786235 1.29 0.257theta=lambda = 0 19.791724 1.28 0.259theta=lambda = 1 18.132125 4.59 0.032---------------------------------------------------------------. regress lg_cost ln_in3 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 1, 74) = 319.49 Model | 11.9217478 1 11.9217478 Prob > F = 0.0000 Residual | 2.76128207 74 .037314623 R-squared = 0.8119-------------+------------------------------ Adj R-squared = 0.8094 Total | 14.6830298 75 .195773731 Root MSE = .19317------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- ln_in3 | .9476271 .053016 17.87 0.000 .8419904 1.053264 _cons | -.835674 .2225597 -3.75 0.000 -1.279134 -.3922142------------------------------------------------------------------------------. estat hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 1.29 Prob > chi2 = 0.2564. estat ovtestRamsey RESET test using powers of the fitted values of lg_cost Ho: model has no omitted variables F(3, 71) = 0.75 Prob > F = 0.5243. mvtest normality lg_cost ln_in3Test for multivariate normality Doornik-Hansen chi2(4) = 5.716 Prob>chi2 = 0.2214. predict uhat,residualslast estimates not foundr(301);. regress lg_cost ln_in3 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 1, 74) = 319.49 Model | 11.9217478 1 11.9217478 Prob > F = 0.0000 Residual | 2.76128207 74 .037314623 R-squared = 0.8119-------------+------------------------------ Adj R-squared = 0.8094 Total | 14.6830298 75 .195773731 Root MSE = .19317------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- ln_in3 | .9476271 .053016 17.87 0.000 .8419904 1.053264 _cons | -.835674 .2225597 -3.75 0.000 -1.279134 -.3922142------------------------------------------------------------------------------. predict uhat,residuals. pnorm uhat. mvtest normality uhatTest for multivariate normality Doornik-Hansen chi2(2) = 1.299 Prob>chi2 = 0.5222. regress lg_cost in1 in4 in5 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 3, 72) = 94.69 Model | 11.7140998 3 3.90469994 Prob > F = 0.0000 Residual | 2.96893003 72 .041235139 R-squared = 0.7978-------------+------------------------------ Adj R-squared = 0.7894 Total | 14.6830298 75 .195773731 Root MSE = .20306------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- in1 | .0983293 .0415578 2.37 0.021 .0154854 .1811732 in4 | .0115103 .0022651 5.08 0.000 .0069948 .0160257 in5 | .0163053 .0046205 3.53 0.001 .0070944 .0255162 _cons | 2.169695 .0682718 31.78 0.000 2.033597 2.305792------------------------------------------------------------------------------. predict uhat,residualsuhat already definedr(110);. predict uhat1,residuals. estat hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 1.33 Prob > chi2 = 0.2485. estat ovtestRamsey RESET test using powers of the fitted values of lg_cost Ho: model has no omitted variables F(3, 69) = 2.83 Prob > F = 0.0446. pnorm uhat1. mvtest normality uhat1Test for multivariate normality Doornik-Hansen chi2(2) = 0.200 Prob>chi2 = 0.9050. regress lg_cost in1 in3 Source | SS df MS Number of obs = 76-------------+------------------------------ F( 2, 73) = 149.72 Model | 11.805049 2 5.90252452 Prob > F = 0.0000 Residual | 2.87798082 73 .039424395 R-squared = 0.8040-------------+------------------------------ Adj R-squared = 0.7986 Total | 14.6830298 75 .195773731 Root MSE = .19856------------------------------------------------------------------------------ lg_cost | Coef. Std. Err. t P>|t| [95% Conf. Interval]-------------+---------------------------------------------------------------- in1 | .0773559 .034649 2.23 0.029 .0083006 .1464113 in3 | .0107338 .0012631 8.50 0.000 .0082164 .0132512 _cons | 2.155491 .0616769 34.95 0.000 2.032569 2.278413------------------------------------------------------------------------------. predict uhat2,residuals. estat hettestBreusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: fitted values of lg_cost chi2(1) = 0.11 Prob > chi2 = 0.7388. estat ovtestRamsey RESET test using powers of the fitted values of lg_cost Ho: model has no omitted variables F(3, 70) = 1.74 Prob > F = 0.1661. pnorm uhat2. mvtest normality uhat2Test for multivariate normality Doornik-Hansen chi2(2) = 1.625
1. Магнус Я.Р., Катышев П.К., Пересецкий А.А. Эконометрика. Начальный курс. М.: Дело, 2000.
2. Доугерти К. Введение в эконометрику. М.:ИНФРА-М, 2001. – 402 с.
3. Тихомиров Н.П., Дорохина Е.Ю. Эконометрика: Учебник. М.: Издательство «Экзамен», 2007. – 512 с.
Построение и анализ на чувствительность моделей задач линейного программирования
Лабораторная работа№. 1
ПОСТРОЕНИЕ И АНАЛИЗ НА ЧУВСТВИТЕЛЬНОСТЬ МОДЕЛЕЙ ЗАДАЧ ЛИНЕЙНОГО ПРОГРАММИРОВАНИЯ
Цель: научиться определять оптимальный план производства (приобретения) продукции с учетом ограниченного обеспечения ресурсами различного типа; освоить методику и технологию поиска оптимального решения задач линейного программирования (ЗЛП) с помощью ЭВМ;, чтобы приобрести практический опыт анализа оптимального решения ЗЛП на чувствительность.
Вариант 1. Для изготовления обуви четырех моделей на фабрике используются два типа кожи. Ресурсы рабочей силы и материала, затраты труда и материала для изготовления каждой пары обуви, но также прибыль от реализации единицы продукции приведены в таблице. Составить план выпуска обуви по ассортименту, максимизирующий прибыль.
| Ресурсов | Количество ресурсов | Затраты ресурсов на одну пару обуви модели | |||
| № 1 | № 2 | № 3 | № 4 | ||
| время Работы, человеко-ч Кожа 1-го сорта Кожа 2-го сорта | 1000 500 1200 | 1 2 0 | 2 1 1 | 2 0 4 | 1 0 1 |
| Прибыль, ден. блок | 2 | 40 | 10 | 15 | |
X1 – количество обуви модели№. 1, выпускаемое фабрикой;
Х2 – количество обуви модели№. 2, выпускаемое фабрикой;
X3 – количество обуви модели№. 3, выпускаемое фабрикой;
X4 – количество обуви модели№. 4, выпускаемое фабрикой.
F = 2*X1 40*X2 10*X3 15*X4 => max - целевая функция
Ограничения ресурсов:
X1 2*X2 2*X3 X4 ≤ 1000
2*X1 X2 ≤ 500
X2 4*X3 X4 ≤ 1200
X1, X2 ≥ 0
Таблица 1.1.