Classifying patterns with missing values using Multi-Task Learning perceptrons

Datasets with missing values are frequent in real-world classification problems. It seems obvious that imputation of missing values can be considered as a series of secondary tasks, while classification is the main purpose of any machine dealing with these datasets. Consequently, Multi-Task Learning (MTL) schemes offer an interesting alternative approach to solve missing data problems. In this paper, we propose an MTL-based method for training and operating a modified Multi-Layer Perceptron (MLP) architecture to work in incomplete data contexts. The proposed approach achieves a balance between both classification and imputation by exploiting the advantages of MTL. Extensive experimental comparisons with well-known imputation algorithms show that this approach provides excellent results. The method is never worse than the traditional algorithms – an important robustness property – and, also, it clearly outperforms them in several problems.     

Improving deep learning performance with missing values via deletion and compensation

Neural Computing and Applications - Missing values in a dataset is one of the most common difficulties in real applications. Many different techniques based on machine learning have been proposed...     

Automatic modelling and simulation of daily global solar radiation series

This paper presents a new model to generate simulated daily global solar radiation (DGSR) sequences. A DGSR value is the product of two factors: a seasonal low-frequency component, principally due to the sun's periodic movement, and a random component due to rapid fluctuations of the atmospheric environment. Methods are provided to automatically separate and estimate both components from the available (generally short) DGSR records. Hence, other series describing the sun-position evolution or any kind of atmospheric conditions are not necessary. An illustrative example is included which results in a good global agreement between simulated and original sequences.     

Compensation for nonlinearities in third-generation mobile systems

A technique for compensating for nonlinear distortion in third-generation (3G) communication systems is presented. In these communication systems, the power-efficient amplification introduces both a widening of the transmitted pulse and nonlinear intersymbol interference. Novel structures are presented which are able to reduce simultaneously both unwanted effects. Examples are presented which show the performance of the compensators in typical digital channels     

Comments on and Extensions on Wolf's Signal-to-Channel Noise Formulas for Delta-Modulated Systems

The channel noise effects on linear delta modulation (LDM) systems have not yet been adequately analyzed. This paper presents a new and general formulation of these effects, based on the theoretical work by Wolf [1]. A comparative discussion of our formulas with previous results is also included. Finally, the application of our methods and the validity of our comments are illustrated by some numerical examples.     

Local estimation of posterior class probabilities to minimize classification errors

Decision theory shows that the optimal decision is a function of the posterior class probabilities. More specifically, in binary classification, the optimal decision is based on the comparison of the posterior probabilities with some threshold. Therefore, the most accurate estimates of the posterior probabilities are required near these decision thresholds. This paper discusses the design of objective functions that provide more accurate estimates of the probability values, taking into account the characteristics of each decision problem. We propose learning algorithms based on the stochastic gradient minimization of these loss functions. We show that the performance of the classifier is improved when these algorithms behave like sample selectors: samples near the decision boundary are the most relevant during learning.     

Cost functions to estimate a posteriori probabilities in multiclassproblems

The problem of designing cost functions to estimate a posteriori probabilities in multiclass problems is addressed. We establish necessary and sufficient conditions that these costs must satisfy in one-class one-output networks whose outputs are consistent with probability laws. We focus our attention on a particular subset of the corresponding cost functions which verify two common properties: symmetry and separability (well-known cost functions, such as the quadratic cost or the cross entropy are particular cases in this subset). Finally, we present a universal stochastic gradient learning rule for single-layer networks, in the sense of minimizing a general version of these cost functions for a wide family of nonlinear activation functions.     

Support vector method for robust ARMA system identification

This paper presents a new approach to auto-regressive and moving average (ARMA) modeling based on the support vector method (SVM) for identification applications. A statistical analysis of the characteristics of the proposed method is carried out. An analytical relationship between residuals and SVM-ARMA coefficients allows the linking of the fundamentals of SVM with several classical system identification methods. Additionally, the effect of outliers can be cancelled. Application examples show the performance of SVM-ARMA algorithm when it is compared with other system identification methods.     

An adaptive beamforming technique based on cyclostationary signalproperties

A new approach to adaptive beamforming is presented. The method is based on the property of cyclostationary signals to generate spectral lines when they pass through certain nonlinear transformations. The beamformer coefficients are selected according to a new optimization objective, which consists on minimizing the mean square error between the array output after the nonlinearity and a complex exponential. This approach optimally extracts any signal that generates a spectral line at the same frequency as the reference complex exponential. A gradient-based algorithm is derived to compute the optimum weights. Since the proposed cost function is a nonconvex function of the array coefficients, minima are analyzed for the three most common types of perturbations found in communications: Gaussian noise, multiple interferences, and multipath propagation. It is demonstrated via analysis and simulations that minima correspond to points where output noise power is minimized, interferences are canceled, and intersymbol interference is removed, i.e., the beamformer eliminates the distortion introduced by the radiocommunication channel     

Pattern classification with missing data: a review

Pattern classification has been successfully applied in many problem domains, such as biometric recognition, document classification or medical diagnosis. Missing or unknown data are a common drawback that pattern recognition techniques need to deal with when solving real-life classification tasks. Machine learning approaches and methods imported from statistical learning theory have been most intensively studied and used in this subject. The aim of this work is to analyze the missing data problem in pattern classification tasks, and to summarize and compare some of the well-known methods used for handling missing values.