Financial decision support using neural networks and support vector machines

Abstract

Bankruptcy prediction and credit scoring are the two important problems facing financial decision support. The multilayer perceptron (MLP) network has shown its applicability to these problems and its performance is usually superior to those of other traditional statistical models. Support vector machines (SVMs) are the core machine learning techniques and have been used to compare with MLP as the benchmark. However, the performance of SVMs is not fully understood in the literature because an insufficient number of data sets is considered and different kernel functions are used to train the SVMs. In this paper, four public data sets are used. In particular, three different sizes of training and testing data in each of the four data sets are considered (i.e. 3:7, 1:1 and 7:3) in order to examine and fully understand the performance of SVMs. For SVM model construction, the linear, radial basis function and polynomial kernel functions are used to construct the SVMs. Using MLP as the benchmark, the SVM classifier only performs better in one of the four data sets. On the other hand, the prediction results of the MLP and SVM classifiers are not significantly different for the three different sizes of training and testing data.

Global Research on Artificial Intelligence from 1990–2014: Spatially-Explicit Bibliometric Analysis

“…Another important keyword is ANN, which has high frequencies of co-occurrence with "ES" and "AI", "prediction", "GA", "fuzzy logic", "SVM", and "CI", showing that ANN is the essential and important algorithm in AI research. By using new methods, in recent years, many scholars have been intent to improve ANN and enhance the efficiency of ANN by coupling with other algorithm, such as GA, and fuzzy logic [54][55][56][57][58][59]. In addition, the close relationship of AI with ANN and ES indicates that many scholars are dedicated to related research and published numerous articles.…”

Measuring effectiveness of a dynamic artificial neural network algorithm for classification problems

“…First, the feature space of a classification problem is not assumed to be linearly separable. Rather, a nonlinear mapping function (also called a kernel function) is used to represent the data in higher dimensions where the boundary between classes is assumed to be linear (Duda, Hart, & Stork, 2001;Muller et al, 2001;Tsai, 2008). Second, the boundary is represented by support vectors instead of a single boundary.…”

“…In Tsai (2008), Tsai examined the effectiveness of neural networks and support vector machines algorithms in financial decisions. Tsai tested several data sets involving credit decisions and bankruptcy.…”

Possibilistic Clustering for Crisis Prediction: Systemic Risk States and Membership Degrees

“…The target variable can be continuous (typically a general prediction problem and usually approached by a type of regression model) or categorical (classification problem). The most widely used methods belonging to this group include decision trees [29], neural networks [19] and support vector machines [26].…”