Dutta, S., Shekhar S. and Wong, W.Y., "Decision support in non--conservative domains: Generalization with neural networks", Decision Support Systems, 11(5), June 1994, pp. 527--544.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Sept 30, 1997. Accepted: April 16, 1998. Corresponding author 1 1. Introduction Successful applications of neural network techniques in business research have been widely reported in the literature. Neural networks have been used to predict bank failures [20] to perform bond rating [4], to analyse financial data [2] to assess the financial health of savings and loan associations [15] to select forecasting models [19] and to assess the need for end users in the planning of the development part of information systems [8] The most attractive feature of neural networks that ....

Dutta, S., Shekhar S. and Wong, W.Y., "Decision support in non--conservative domains: Generalization with neural networks", Decision Support Systems, 11(5), June 1994, pp. 527--544.

....ideally suited for this type of forecasting problem with many inputs and a single output as it has a direct structure and can be easily trained. Also, it has proved to be efficient in a number of applications, including predicting sales [6] forecasting prices [9] assigning corporate bond ratings [4], predicting thrift failures [16] dynamic modeling of stock returns [13] predicting stock returns [5] and predicting bankruptsy [8] The number of input and output units was defined by the problem. We developed neural network models using both 18 input units and 26 input units. One output for ....

Dutta C. G. and Shekhar W. Y. Decision support in non-conservative domains: generalisation with neural networks. Decision Support Systems, 11:527--544, 1994.

....of the unrelatedness and non linearity of marketing data. A study was carried out by Proctor [6] on the role of neural networks in marketing. The strength of a neural network is its ability to obtain the relationships of non linearly dependent variables. This was further emphasized by Dutta et al. [7], who incorporated neural networks into a Decision Support System (DSS) In a separate work, Venugopal and Baets [8] also proposed using neural networks in retail sales forecasting, direct marketing and target marketing. Compared with other traditional statistical methods, neural networks require ....

....there are no learning facilities although new situations arise and the results need to be estimated frequently. Neural Networks do not suffer from the limitations of regression models and have been proven to be able to learn functional relationships from input variables to predict results [3] [7] [10] 11] In our experiments, one hidden layer backpropagation neural networks were used. The number of nodes in a hidden layer ranged from one to two thirds of the number in input layers. The activation function used was a sigmoid function. 3 Background Information In Singapore, consumer ....

S. Dutta, S. Shekhar, and W.Y. Wong, "Decision Support In Non-Conservative Domains: Generalisation With Neural Networks" , Decision Support Systems, 11, 1994.

....ideally suited for this type of forecasting problem with many inputs and a single output as it has a direct structure and can be easily trained. Also, it has proved to be efficient in a number of applications, including predicting sales [6] forecasting prices [1] assigning corporate bond ratings [3], predicting thrift failures [9] dynamic modeling of stock returns [12] predicting stock returns [13] and predicting bankruptsy [15] The number of input and output units was defined by the problem. We developed neural network models using 26 input units. One output for predicting continuous ....

C. G. Dutta and W. Y. Shekhar. Decision support in non-conservative domains: generalisation with neural networks. Decision Support Systems, 11:527--544, 1994.

No context found.

Dutta, S., S. Shekhar and W.Y. Wong,, "Decision Support In Non-Conservative Domains: Generalisation With Neural Networks" , Decision Support Systems, 11, 1994.

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