D. Harrison and D. L. Rubinfeld. Hedonic prices and the demand for clean air. Journal of Environ. Economics and Management, 5:81--102, 1978.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....distribution width for the gth weight group, introduced in chapter 13. 14.1 A real world regression problem: The Boston houseprice data In a final study, the GM RVFL network and the ARD scheme derived in the previous chapter were applied to the Boston house price data of Harrison and Rubinfeld [24]. This is a real world regression problem that has been studied by several authors before and can be used as a benchmark set . For each of the 506 census tracts within the Boston metropolitan area the data give 13 socio economic explanatory variables, as listed in See ....

.... rate per 10,000 PTRATIO pupil teacher ratio by town BLACK proportion of blacks by town LOWSTAT per cent lower status of the population MEDV median value of owner occupied homes in 1000 s Table 14.1: Explanatory variables (above) and target (below) of the Boston housing data set [24]. with standard deviation oe rand ) The networks were trained with the EM algorithm, including the Bayesian regularisation scheme I, described in section 9.4.2. The update of the network parameters thus followed equations (9.83) 9.84) and (9.87) The data were split into two disjoint sets of ....

[Article contains additional citation context not shown here]

Harrison D. and Rubinfeld D.L. (1978): Hedonic prices and the demand for clean air. Journal of Environmental Economics and Management 5, 81-102.

....form are however usually not dictated by economic theory. Other well known examples are labor wages as a function of age and education (see e.g. 11] or so called hedonic price models where the price of a consumer good depends on a bundle of charac teristics for which a valuation exists [9]. Another class of problems where monotonicity constraints often apply are so called selection problems. Consider for example the selection of applicants for a job or a loan on the basis of their characteristics. Because the monotonicity constraint is quite common in practice, many data analysis ....

O. Harrison and D. Rubinfeld. Hedonic prices and the demand for clean air. Journal of Environmental Economics and Management, 53:81-102, 1978.

....There is no proof that such a vector exists but empirically this is usually the case [20] If all support vectors have C = # then the solution is said to be unstable, as the global optimum is not unique. In this case the optimal bias can be calculated by an appeal to the geometry of the hyperplane [16]. SVM learns the optimal separating hyperplane in some feature space, subject to ignoring certain points, which could be interpreted as training misclassifications. The learned hyperplane is an expansion on a subset of the training data, which is known as the support vectors. By use of an ....

....that corresponds to a known classifier such as a Gaussian RBF, sigmoid or polynomial classifier. This is possible since Mercer s theorem states that any positive definite kernel corresponds to an inner product in some feature space. Kernels can also be constructed to incorporate domain knowledge [16,17]. Fig. 1. a) Dataset normal distribution (b) decision boundary formed by SVM using linear kernel Fig. 2. a) Normal target position of a dataset (b) target position of the dataset after smoothing This so called kernel trick gives the SVM great flexibility. With a suitable choice of ....

D. Harrison and D.L. Rubinfeld. Hedonic prices and the demand for clean air. J. Environ. Economics and Management, 5:81-102, 1978.

....in this thesis will be able to be judged in a wider context. At present, however, it is possible to still judge the results in comparison with the existing results in DELVE on the data sets described here. For the regression tasks the only natural data set used is the Boston housing data [78]. The other data sets represent particular versions of two families of data sets: the Kin family and Pumadyn family. These are realistic simulations of the kinematics and dynamics of robot arms. The Kin and Pumadyn data sets span a range of different types of problems: from medium to high noise ....

....The Boston Housing data set Each of the attributes of the Boston data set are shown with its name, whether it is categorical or uniform (c u) its range and its description. the attributes are separated into inputs and a single target. The Boston housing data set was originally published in [78]. The data contains information about housing in the Boston Massachusetts area collected by the United States Census Service. The data is also available in the StatLib archive The data set has been widely used throughout statistical literature as a benchmark. The prototask considered in this ....

[Article contains additional citation context not shown here]

Harrison, D. and Rubinfeld, D. [1978], `Hedonic prices and the demand for clean air', Journal of Environmental Economics & Management 5, 81--102.

....of the new criteria for regression and two class classification. 5. ANEXAMPLEOFREGRESSIONTREES: THE BOSTON HOUSING DATA Following CART, we demonstrate the application of the new splitting criteria on the Boston Housing data. These well known data were originally created by Harrison and Rubinfeld [5], and they were popularized by Belsley, Kuh and Welsch [1] Data files are available from the UC Irvine Machine Learning Repository [6] Harrison and Rubinfeld s main interest in the data was to investigate how air pollution concentration (NOX) a#ects the value of single family homes in the ....

Harrison, R. J., and Rubinfeld, D. L. (1978), "Hedonic Prices and the Demand for Clean Air," Journal of Environmental Economics and Management, 5, 81-102.

....a specified outcome variable, than to cases in other subregions. For a classification tree, the cases within a subregion would belong predominantly to one class. For a regression tree, the cases within a subregion have a small mean squared error. To illustrate, consider the Boston housing data [Harrison and Rubenfeld, 1978; Breiman et al. 1984] consisting of 14 variables measured on 506 census tracts in the Boston area. The outcome of interest is the log median value (MV) of the houses in the tract as a function of the crime rate (CRIM) percent of land zoned for lots (ZN) percent nonretail business (INDUS) ....

....to partition sibpair data on binary splits of covariates to isolate subgroups where genetic models are operating. Once these subgroups are found, within subgroup linkage analyses may be performed. To illustrate the new splitting rule, consider again the Boston housing data. In the original paper [Harrison and Rubenfeld, 1978], the interest was in modeling the effect of air pollution (variable NOX) on log median housing costs within a census tract. Figure 2 shows a linear relationship for all 506 census tracts where an increase in air pollution is associated with a decrease in the median value of the homes. ....

Harrison D, Rubinfeld D. 1978. Hedonic prices and the demand for clean air. J Environ Econ Manage 5:81--102.

....two orders of magnitude. This allows the creation of a large committee of predictors, which was found to lead to an improved generalisation performance. The following study aims at applying the GM RVFL network to a real world benchmark problem, the Boston house price data of Harrison and Rubinfeld [5]. For each of the 506 census tracts within the Boston metropolitan area, the data give 13 socio economic explanatory variables, like crime rate, air pollution, or distance to main employment centres. The target is to predict the median housing price for a given tract. There are several hints that ....

....Com gen according to (38) we followed an idea by Sharkey et al. 19] and subjected the data to different nonlinear transformations. In scheme Tr1, all inputs (except for one binary indicator variable) were subjected to a logarithmic transformation. In scheme Tr2, the transformation described in [5] was applied, where one input variable is squared, and three input variables are replaced by their logarithm. After either transformation, the data were normalised again. A total of 200 networks were trained on the data subjected to Tr1, 200 networks on those subjected to Tr2. The remaining 200 ....

Harrison D. and Rubinfeld D.L. (1978): Hedonic prices and the demand for clean air. Journal of Environmental Economics and Management 5, 81-102.

....DESIGN FOR THE MORTALITY RATE PREDICTION PROBLEM. K IS THE NUMBER OF GAUSSIAN BASIS FUNCTIONS with the DA result. Since the regression function obtained by DA is generally independent of initialization, a single DA run sufficed. First, we considered the Boston home value prediction problem [10]. Here, we use data from 506 homes in the Boston area to predict the median price of each home from 13 features that are believed to have some influence on it. Since the features have different dynamic ranges, we first normalized each one to unit variance. Using the entire data for training, we ....

D. Harrison and D. L. Rubinfeld, "Hedonic prices and the demand for clean air," J. Environ. Econom. Manag. vol. 5, pp. 81--102, 1978.

....regression methods. However, due to the unavailability of sufficient data for proper validation, we simply compare the performance of the two regression models versus model size. One benchmark problem is concerned with predicting the value of homes in the Boston area from a variety of parameters [43]. The training set consists of data from 506 homes. The output in this case is the median price of a home, with the input consisting of a vector of 13 scalar features believed to influence the price. The objective is to minimize the average squared error in price prediction. Since the features ....

D. Harrison and D. L. Rubinfeld, "Hedonic prices and the demand for clean air," J. Environ. Economics and Management, vol. 5, pp. 81--102, 1978.

....di erent classes of models are developed, and used to cluster models into similar groups. Although many of the ideas are general, the focus here will be on linear models and trees, with examples using tree models. To illustrate the problem of multiple models, consider the Boston Housing data (Harrison and Rubinfeld 1978). 506 census tracts in the greater Boston area have 14 recorded characteristics, such as crime rate, proportion of lower status individuals, parent teacher ratios, pollution levels, median house value, and others. The goal is to predict median house value using the other 13 variables as ....

Harrison, D. and Rubinfeld, D.L. (1978) \Hedonic Prices and the Demand for Clean Air", Journal Environmental Economics and Management, 5, 81-102.

....t both global non linearity and global heteroskedasticity. In subsequent sections, we will describe in detail a Bayesian approach for nding and tting treed regressions to data. In particular, in Section 6. 3, we illustrate the performance of our approach on the well known Boston Housing data, Harrison and Rubinfeld (1978). To give the reader an immediate sense of the potential of our approach, we describe here some of what we found. The data consist of 14 characteristics of 506 census tracts in the Boston area. The response is the logged median value of owner occupied homes in each tract, and is listed in Table ....

Harrison, D. and Rubinfeld, D. L. (1978) \Hedonic Prices and the Demand for Clean Air", Journal of Environmental Economic and Management, 5, 81-102.

....The Boston Housing Dataset : This is a well known dataset concerning tracts in Boston. Thirteen (mainly socio economic) variables are available as input for each tract. The target value is the median price for owner occupied houses in the tract. The number of samples in the dataset is 506. See [3] for more information. 5.2 Results To validate that the above described method was correctly implemented we applied the program to the arti cial dataset using various values of , the parameter that controls the contribution of the CT part to the total error function. One would expect that as ....

H. Harrison and D.L. Rubinfeld. Hedonic prices and the demand for clean air. J. Environ. Economics & Management, 5:81-102, 1978.

....rapidly convert the rest of the population into their own image. Figure 8 on the color plate presents a network trained to predict the average market value of houses in a development based on 13 inputs, including property tax rate, student teacher ratios, and distance to major employment centers [3, 7]. This network has the convenient feature that the weights emanating from the last two layers are all positive, so that the sign of the weights emanating from a specific input node gives a direct indication of the influence that node is having on the final output. Looking at input node (1) for ....

D. Harrison, D.L. Rubinfeld. Hedonic prices and the demand for clean air. Journal of Environmental Economics and Management, Vol. 5, pages 81-102, 1978.

....method only includes order p: K(x, y) K p (x, y) 16) In our experiments we used ANOVA decompositions of the latter type, only considering a term of exactly order p. 5 Experiments 5. 1 Data Our experiments were conducted on the Boston housing problem from StatLib at Carnegie Mellon University [HR78] This is a well known data set for testing non linear regression methods. Previous uses include Drucker et al. DBK 97] and Breiman [Bre94] The data set consists of 506 cases in which 12 continuous variables and 1 binary variable determine the median house price in a certain area of Boston ....

D. Harrison and D. L. Rubinfeld. Hedonic prices and the demand for clean air. In J. Environ. Economics & Management, volume 5, pages 81--102, 1978. Supposed original source, data actually from ftp://ftp.ics.uci.com/pub/machine-learning-databases/housing.

....0.25 0.5 0.75 1 39 31 21 11 0 11 21 31 39 Estimate Verification Decision Parameter z fi 0 0.25 0.5 0. 75 1 39 31 21 11 0 11 21 31 39 Estimate Verification Decision Parameter z Australian Credit Card Estimated and verified ff and fi Boston Housing The data set was created by Harrison and Rubinfeld (1978). The data consist of 506 records concerning housing values in the Boston area. Each record is composed of 13 attributes, of which 12 have rational values, while one is Boolean. The median value of the owner occupied houses is used as a threshold to split the entire set of records into two sets. ....

Harrison, D., and Rubinfeld, D. L. (1978), Hedonic prices and the demand of clean air, Journal of Environment Economics and Management, 5, 111-143.

....p = 0:2 rather excludes correct indices. We conclude thus from this table that a 90 percent resolution of variation (i.e. 0:9) shows the best performance and could be recommended for practice. In order to illustrate our method on a real data set we applied it to the house pricing data set of Harrison and Rubinfeld (1978). The vector of explanatory variables X was d = 12 dimensional and included important socioeconomic and geographic variables such as crime rate (X 1 ) percent lower status (X 12 ) and proportion of blacks (X 11 ) For a complete list of variables see Breiman, Friedman, Olshen and Stone (1984, p. ....

Harrison, D. and Rubinfeld, D.L. (1978). Hedonic prices and the demand for clean air. J. Envir. Econ. and Management, 5, 81-102.

....for performing fully principled comparisons between learning methods. In particular it takes into account training set and test cases effects that are overlooked by standard paired tests on single test sets. We will use 6 datasets in our comparison: Boston NOX The Boston housing dataset [18], predicting the nitric oxides concentration from 13 other attributes, 506 cases in all. Boston Price Same as above but predicting the median value of owner occupied homes. Pumadyn 8fh Data from a realistic simulation of the dynamics of a Puma 560 robot arm, predicting the angular accelaration ....

D. Harrison and D. L. Rubinfeld. Hedonic prices and the demand for clean air. J. Environ. Economics and Management, 5:81--102, 1978.

....and 270 records. The StatLog Project employed unequal misclassi cation costs. We use equal costs here because some algorithms do not allow unequal costs. The error rates are estimated using ten fold cross validation. Boston housing (bos) This UCI dataset gives housing values in Boston suburbs [20]. There are three classes, twelve numerical attributes, one binary attribute, and 506 records. Following Loh and Vanichsetakul (1988) the classes are created from the attribute median value of owner occupied homes as follows: class = 1 if log(median value) 9:84, class = 2 if 9.84 log(median ....

D. Harrison and D. L. Rubinfeld. Hedonic prices and the demand for clean air. Journal of Environmental Economics and Management, 5:81-102, 1978.

No context found.

D. Harrison and D. L. Rubinfeld. Hedonic prices and the demand for clean air. Journal of Environ. Economics and Management, 5:81--102, 1978.

No context found.

D. Harrison, D.L. Rubinfeld, "Hedonic prices and the demand for clean air", Journal of Environmental Economics and Management, Vol. 5, pp. 81-102, 1978.

No context found.

D. Harrison, D. Rubinfeld, Hedonic prices and the demand for clean air, J. Environ. Economics & Management 5 (1978) 81--102.

No context found.

Harrison, D. and Rubinfeld, D. L. (1978). Hedonic prices and the demand for clean air. J. Environ. Economics & Management 5, 81-102

No context found.

Harrison, R. J., and Rubinfeld, D. L. (1978), "Hedonic Prices and the Demand for Clean Air," Journal of Environmental Economics and Management, 5, 81-102.

No context found.

Harrison, D. and Rubinfeld, D.L. (1978). Hedonic prices and the demand for clean air. J. Environ. Economics and Management, 5, 81-102.

No context found.

Harrison, D. and Rubinfeld, D. L. (1978) Hedonic prices and the demand for clean air. J. Environ. Economics & Management, 5, 81--102.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC