Alfredo Vellido

Neural networks in business: a survey of applications (1992–1998)

Abstract During the last decade, neural networks have established themselves as a theoretically sound alternative to traditional statistical models, and a large body of research on their application to business has been produced. The comprehensive range of business and financial applications is such that a focus is required for an in-depth analysis, therefore this review addresses applications related to management, marketing and decision making. Also, given that previous reviews have dealt with earlier publications, the time span of the review is limited to the period 1992–1998. The presentation is centred on summary tables with links between them. These tables classify the studies according to their application areas, the main contributions rendered by the use of neural networks, and the alleged advantages and disadvantages of this, as well as the journal of publication. Further information on the neural network models, other statistical methods against which they have been compared, and features of the analysed data are also provided. The more controversial issues concerning real-world applications of neural networks are discussed as a part of a critical analysis. Many of the studies are shown to be first attempts to apply these new techniques to established areas of research, whereas only a few tackle real-world cases. Although still regarded as a novel methodology, neural networks are shown to have matured to the point of offering real practical benefits in many of their applications.

Applying Data Mining Techniques to e-Learning Problems

This chapter aims to provide an up-to-date snapshot of the current state of research and applications of Data Mining methods in e-learning. The cross-fertilization of both areas is still in its infancy, and even academic references are scarce on the ground, although some leading education-related publications are already beginning to pay attention to this new field. In order to offer a reasonable organization of the available bibliographic information according to different criteria, firstly, and from the Data Mining practitioner point of view, references are organized according to the type of modeling techniques used, which include: Neural Networks, Genetic Algorithms, Clustering and Visualization Methods, Fuzzy Logic, Intelligent agents, and Inductive Reasoning, amongst others. From the same point of view, the information is organized according to the type of Data Mining problem dealt with: clustering, classification, prediction, etc.

Quantitative characterization and prediction of on-line purchasing behavior: a latent variable approach

Abstract: Realizing the full potential of the on-line consumer market requires careful identification of customer needs and expectations. As research on Internet consumer behavior is still in its infancy, a quantitative framework to characterize user profiles for e-commerce has not yet been established. This study proposes a quantitative framework that uses factor analysis to identify latent factor descriptors of Internet users’ opinions on Web vendors and on-line shopping. Predictive models based on logistic discrimination and neural networks then select the factors most predictive of the propensity to buy on-line and classify Internet users accordingly. The application of this framework shows that the obtained latent factors agree in general with the major indicators identified in previous qualitative research. A small subset of the obtained factors is shown to retain the predictive power of the whole set. Neural networks are found to perform only marginally better than logistic discrimination in the task of classification.

Missing data imputation through GTM as a mixture of t-distributions

The Generative Topographic Mapping (GTM) was originally conceived as a probabilistic alternative to the well-known, neural network-inspired, Self-Organizing Maps. The GTM can also be interpreted as a constrained mixture of distribution models. In recent years, much attention has been directed towards Student t-distributions as an alternative to Gaussians in mixture models due to their robustness towards outliers. In this paper, the GTM is redefined as a constrained mixture of t-distributions: the t-GTM, and the Expectation-Maximization algorithm that is used to fit the model to the data is modified to carry out missing data imputation. Several experiments show that the t-GTM successfully detects outliers, while minimizing their impact on the estimation of the model parameters. It is also shown that the t-GTM provides an overall more accurate imputation of missing values than the standard Gaussian GTM.

Business Applications of Neural Networks:The State-of-the-Art of Real-World Applications

Neural networks are increasingly being used in real-world business applications and, in some cases, such as fraud detection, they have already become the method of choice. Their use for risk assessment is also growing and they have been employed to visualise complex databases for marketing segmentation. This boom in applications covers a wide range of business interests — from finance management, through forecasting, to production. The combination of statistical, neural and fuzzy methods now enables direct quantitative studies to be carried out without the need for rocket-science expertise.

Classification of human brain tumours from MRS data using Discrete Wavelet Transform and Bayesian Neural Networks

The diagnosis of brain tumours is an extremely sensitive and complex clinical task that must rely upon information gathered through non-invasive techniques. One such technique is Magnetic Resonance Spectroscopy. In this task, radiology experts are likely to benefit from the support of computer-based systems built around robust classification processes. In this paper, a Discrete Wavelet Transform procedure was applied to the pre-processing of spectra corresponding to several brain tumour pathologies. This procedure does not alleviate the high dimensionality of the data by itself. For this reason, dimensionality reduction was subsequently implemented using Moving Window with Variance Analysis for feature selection or Principal Component Analysis for feature extraction. The combined method yielded very encouraging results in terms of diagnostic discriminatory binary classification using Bayesian Neural Networks. In most cases, the classification accuracy improved on previously reported results.

Selective smoothing of the generative topographic mapping

Generative topographic mapping is a nonlinear latent variable model introduced by Bishop et al. as a probabilistic reformulation of self-organizing maps. The complexity of this model is mostly determined by the number and form of basis functions generating the nonlinear mapping from latent space to data space, but it can be further controlled by adding a regularization term to increase the stiffness of the mapping and avoid data over-fitting. In this paper, we improve the map smoothing by introducing multiple regularization terms, one associated with each of the basis functions. A similar technique to that of automatic relevance determination, our selective map smoothing locally controls the stiffness of the mapping depending on length scales of the underlying manifold, while optimizing the effective number of active basis functions.

Advances in clustering and visualization of time series using GTM through time

Most of the existing research on multivariate time series concerns supervised forecasting problems. In comparison, little research has been devoted to their exploration through unsupervised clustering and visualization. In this paper, the capabilities of Generative Topographic Mapping Through Time, a model with foundations in probability theory, that performs simultaneous time series clustering and visualization, are assessed in detail. Focus is placed on the visualization of the evolution of signal regimes and the exploration of sudden transitions, for which a novel identification index is defined. The interpretability of time series clustering results may become extremely difficult, even in exploratory visualization, for high dimensional datasets. Here, we define and test an unsupervised time series relevance determination method, fully integrated in the Generative Topographic Mapping Through Time model, that can be used as a basis for time series selection. This method should ease the interpretation of time series clustering results.

Feature and model selection with discriminatory visualization for diagnostic classification of brain tumors

Machine Learning (ML) and related methods have of late made significant contributions to solving multidisciplinary problems in the field of oncology diagnosis. Human brain tumor diagnosis, in particular, often relies on the use of non-invasive techniques such as Magnetic Resonance Imaging (MRI) and Spectroscopy (MRS). In this paper, MRS data of human brain tumors are analyzed in detail. The high dimensionality of the MR spectra makes difficult both their classification and the interpretation of the obtained results, thus limiting their usability in practical medical settings. The use of dimensionality reduction techniques is therefore advisable. In this work, we apply feature selection methods and several off-the-shelf classifiers on various ^1H-MRS modalities: long and short echo times and an ad hoc combination of both. The introduction of bootstrap resampling techniques permits the obtention of mean performance estimates and their variability. Our experimental findings indicate that the feature selection process enhances the classification performance compared to using the full set of features. We also show that the use of combined information from the different echo times is a better strategy for small numbers of spectral frequencies; however, the use of ever greater numbers of short echo time frequencies permits the obtention of many models with similar performance. The final induced models offer very attractive solutions both in terms of prediction accuracy and number of involved spectral frequencies, which are also amenable to metabolic interpretation. A linear dimensionality-reduction technique that preserves class discrimination capabilities is used for visualizing the data corresponding to the selected frequencies.

Bias reduction in skewed binary classification with Bayesian neural networks

The Bayesian evidence framework has become a standard of good practice for neural network estimation of class conditional probabilities. In this approach the conditional probability is marginalised over the distribution of network weights, which is usually approximated by an analytical expression that moderates the network output towards the midrange. In this paper, it is shown that the network calibration is considerably improved by marginalising to the prior distribution. Moreover, marginalisation to the midrange can seriously bias the estimates of the conditional probabilities calculated from the evidence framework. This is especially the case in the modelling of censored data.