Robust loss functions for boosting

Boosting is known as a gradient descent algorithm over loss functions. It is often pointed out that the typical boosting algorithm, Adaboost, is highly affected by outliers. In this letter, loss functions for robust boosting are studied. Based on the concept of robust statistics, we propose a transformation of loss functions that makes boosting algorithms robust against extreme outliers. Next, the truncation of loss functions is applied to contamination models that describe the occurrence of mislabels near decision boundaries. Numerical experiments illustrate that the proposed loss functions derived from the contamination models are useful for handling highly noisy data in comparison with other loss functions.     

Selectivity and stability via dendritic nonlinearity

Inspired by recent studies regarding dendritic computation, we constructed a recurrent neural network model incorporating dendritic lateral inhibition. Our model consists of an input layer and a neuron layer that includes excitatory cells and an inhibitory cell; this inhibitory cell is activated by the pooled activities of all the excitatory cells, and it in turn inhibits each dendritic branch of the excitatory cells that receive excitations from the input layer. Dendritic nonlinear operation consisting of branch-specifically rectified inhibition and saturation is described by imposing nonlinear transfer functions before summation over the branches. In this model with sufficiently strong recurrent excitation, on transiently presenting a stimulus that has a high correlation with feed- forward connections of one of the excitatory cells, the corresponding cell becomes highly active, and the activity is sustained after the stimulus is turned off, whereas all the other excitatory cells continue to have low activities. But on transiently presenting a stimulus that does not have high correlations with feedforward connections of any of the excitatory cells, all the excitatory cells continue to have low activities. Interestingly, such stimulus-selective sustained response is preserved for a wide range of stimulus intensity. We derive an analytical formulation of the model in the limit where individual excitatory cells have an infinite number of dendritic branches and prove the existence of an equilibrium point corresponding to such a balanced low-level activity state as observed in the simulations, whose stability depends solely on the signal-to-noise ratio of the stimulus. We propose this model as a model of stimulus selectivity equipped with self-sustainability and intensity-invariance simultaneously, which was difficult in the conventional competitive neural networks with a similar degree of complexity in their network architecture. We discuss the biological relevance of the model in a general framework of computational neuroscience.     

Patient views of adverse events: comparisons of self-reported healthcare staff attitudes with disclosure of accident information

In the present paper, we report results of surveys in 2003 in Japan and Denmark about patients' views about adverse events, focusing on the actions of healthcare staff involved in a medical accident. Results show that patients were more likely to indicate negative expectations to a doctor's reactions after a medical accident when asked in general terms than when asked in relation to concrete case stories. When asked in general terms, 66% (47%) of Japanese (Danish) respondents expected that doctors sometimes hold back on providing information to patients about a medical accident, while 37% (7%) did so when asked about a concrete, mild-outcome case. We examine some possible reasons for the relatively high level of distrust of Japanese patients, and we discuss whether the seemingly lower level of disclosure in Japan than in Denmark and the negative stories in the Japanese press may have an impact. We also suggest some implications for introducing a patient-centred or customer-centred approach to risk management in healthcare and other domains.     

A new learning method using prior information of neural networks

In this paper, we present a new learning method using prior information for three-layer neural networks. Usually when neural networks are used for identification of systems, all of their weights are trained independently, without considering interrelated weights values. Thus, the training results are usually not good. The reason for this in that each parameter has its influence on others during learning. To overcome this problem, we first give an exact mathematical equation that describes the relation between weight values given a set of data conveying prior information. The we present a new learning method that trains part of the weights and calculates the others using these exact mathematical equations. This method often a priori keeps the given mathematical structure exactly the same during learning; in other words, training is done so that the network follows a predetermined trajectory. Numerical computer simulation results are provided to support this approach. This work was presented, in part, at the Fourth International Symposium on Artificial Life and Robotics, Oita, Japan, January 19–22, 1999.     

Rapid reinforcement for fusion mass-spliced fibers using low-power induction heating

A rapid reinforcement method using low-power induction heating is developed. The reinforcement of a mass-spliced fiber ribbon unit comprising five graded-index multimode fibers was completed in 30 sec with good performance by supplying 30 W of electric power.     

Flow and deformation of fresh concrete

The results of studies made from the standpoint of rheology on methods of predicting flows and deformations of fresh concrete and mortar as basic research for rationalization of concrete construction are described. This paper contains as follows:

1. method of measuring viscosity of fresh concrete and mortar;
2. method of estimating deformation of fresh concrete due to its own weight;
3. proposal of the “Inclined Pipe Test Method” as a new method of testing consistency of grout mortar.

     

Further investigation on the thermal decomposition of aluminium dialkylamides

Both dimethylamino- (I) and diethylamino-alane dimers (II), [(R₂N)₂AlH]₂ (I, R=Me; II, R=Et) decompose above 800° C under a few hundred Pa of hydrogen to dark greyish, hard (Vickers hardness larger than 2000), oxidation-resistive and oxidation-protective deposits which are tentatively identified as Al₅C₃N. The deposits are stable to moisture and diluted hydrochloric acid in contrast to those obtained below 800° C, but they easily dissolve in sodium hydroxide solution at room temperature evolving a gas. The deposits on a stainless steel substrate adhere strongly to the substrate and remain so on rapid heating and cooling. The electrical resistivity of the deposits is in the range 10² to 10⁴ cm.     

High rate nitrogen removal by the CANON process at ambient temperature

Completely autotrophic nitrogen removal over nitrite (CANON) is a cost-effective nitrogen removal process. Implementation of the CANON process relies on the cooperation of ammonium-oxidizing and Anammox bacteria, as well as the inhibition of nitrite-oxidizing bacteria. Strict limitations on dissolved oxygen (DO) concentration in the reactor, and the addition of sufficient inorganic carbon in the influent, were adopted as the main operational strategies. The reactor was fed with synthetic inorganic wastewater composed mainly of NH(4)(+)-N, and operated for 106 days. Stable nitrogen removal rates (NRR) of around 1.4 kg N m(-3) d(-1) were obtained at ambient temperature. Morphological characteristics and analysis of bacterial community confirmed the formation of functional outer aerobic and inner anaerobic granular sludge, providing evidence of stable nitrogen removal.     

Multiple kernel learning with gaussianity measures

Kernel methods are known to be effective for nonlinear multivariate analysis. One of the main issues in the practical use of kernel methods is the selection of kernel. There have been a lot of studies on kernel selection and kernel learning. Multiple kernel learning (MKL) is one of the promising kernel optimization approaches. Kernel methods are applied to various classifiers including Fisher discriminant analysis (FDA). FDA gives the Bayes optimal classification axis if the data distribution of each class in the feature space is a gaussian with a shared covariance structure. Based on this fact, an MKL framework based on the notion of gaussianity is proposed. As a concrete implementation, an empirical characteristic function is adopted to measure gaussianity in the feature space associated with a convex combination of kernel functions, and two MKL algorithms are derived. From experimental results on some data sets, we show that the proposed kernel learning followed by FDA offers strong classification power.