Kazuyuki Murase

MWMOTE--Majority Weighted Minority Oversampling Technique for Imbalanced Data Set Learning

Imbalanced learning problems contain an unequal distribution of data samples among different classes and pose a challenge to any classifier as it becomes hard to learn the minority class samples. Synthetic oversampling methods address this problem by generating the synthetic minority class samples to balance the distribution between the samples of the majority and minority classes. This paper identifies that most of the existing oversampling methods may generate the wrong synthetic minority samples in some scenarios and make learning tasks harder. To this end, a new method, called Majority Weighted Minority Oversampling TEchnique (MWMOTE), is presented for efficiently handling imbalanced learning problems. MWMOTE first identifies the hard-to-learn informative minority class samples and assigns them weights according to their euclidean distance from the nearest majority class samples. It then generates the synthetic samples from the weighted informative minority class samples using a clustering approach. This is done in such a way that all the generated samples lie inside some minority class cluster. MWMOTE has been evaluated extensively on four artificial and 20 real-world data sets. The simulation results show that our method is better than or comparable with some other existing methods in terms of various assessment metrics, such as geometric mean (G-mean) and area under the receiver operating curve (ROC), usually known as area under curve (AUC).

Contribution of microglia and astrocytes to the central sensitization, inflammatory and neuropathic pain in the juvenile rat

BackgroundThe development of pain after peripheral nerve and tissue injury involves not only neuronal pathways but also immune cells and glia. Central sensitization is thought to be a mechanism for such persistent pain, and ATP involves in the process. We examined the contribution of glia to neuronal excitation in the juvenile rat spinal dorsal horn which is subjected to neuropathic and inflammatory pain.ResultsIn rats subjected to neuropathic pain, immunoreactivity for the microglial marker OX42 was markedly increased. In contrast, in rats subjected to inflammatory pain, immunoreactivity for the astrocyte marker glial fibrillary acidic protein was increased slightly. Optically-recorded neuronal excitation induced by single-pulse stimulation to the dorsal root was augmented in rats subjected to neuropathic and inflammatory pain compared to control rats. The bath application of a glial inhibitor minocycline and a p38 mitogen-activated protein kinase inhibitor SB203580 inhibited the neuronal excitation in rats subjected to neuropathic pain. A specific P2X1,2,3,4 antagonist TNP-ATP largely inhibited the neuronal excitation only in rats subjected to neuropathic pain rats. In contrast, an astroglial toxin L-alpha-aminoadipate, a gap junction blocker carbenoxolone and c-Jun N-terminal kinase inhibitor SP600125 inhibited the neuronal excitation only in rats subjected to inflammatory pain. A greater number of cells in spinal cord slices from rats subjected to neuropathic pain showed Ca2+ signaling in response to puff application of ATP. This Ca2+ signaling was inhibited by minocycline and TNP-ATP.ConclusionsThese results directly support the notion that microglia is more involved in neuropathic pain and astrocyte in inflammatory pain.

Evolutionary Path Control Strategy for Solving Many-Objective Optimization Problem

The number of objectives in many-objective optimization problems (MaOPs) is typically high and evolutionary algorithms face severe difficulties in solving such problems. In this paper, we propose a new scalable evolutionary algorithm, called evolutionary path control strategy (EPCS), for solving MaOPs. The central component of our algorithm is the use of a reference vector that helps simultaneously minimizing all the objectives of an MaOP. In doing so, EPCS employs a new fitness assignment strategy for survival selection. This strategy consists of two procedures and our algorithm applies them sequentially. It encourages a population of solutions to follow a certain path reaching toward the Pareto optimal front. The essence of our strategy is that it reduces the number of nondominated solutions to increase selection pressure in evolution. Furthermore, unlike previous work, EPCS is able to apply the classical Pareto-dominance relation with the new fitness assignment strategy. Our algorithm has been tested extensively on several scalable test problems, namely five DTLZ problems with 5 to 40 objectives and six WFG problems with 2 to 13 objectives. Furthermore, the algorithm has been tested on six CEC09 problems having 2 or 3 objectives. The experimental results show that EPCS is capable of finding better solutions compared to other existing algorithms for problems with an increasing number of objectives.

Astrocytes are involved in long-term facilitation of neuronal excitation in the anterior cingulate cortex of mice with inflammatory pain.

Summary Astrocytes are involved in long‐term facilitation of neuronal excitation in the anterior cingulate cortex of mice with inflammatory pain. Hiroshi Ikeda, Keiichi Mochizuki, Kazuyuki Murase. Activated astrocytes in the anterior cingulate cortex play a crucial role in the long‐term potentiation of and the development of negative emotions in conditions of inflammatory pain. Abstract Neuronal plasticity in the pain‐processing pathway is thought to be a mechanism underlying pain hypersensitivity and negative emotions occurring during a pain state. Recent evidence suggests that the activation of astrocytes in the anterior cingulate cortex (ACC) contributes to the development of negative emotions during pain hypersensitivity after peripheral inflammation. However, it is unknown whether these activated astrocytes contribute to neuronal plasticity in the ACC. In this study, by using optical imaging with voltage‐ and Ca2+‐sensitive dyes, we examined the long‐term facilitation of neuronal excitation induced by high‐frequency conditioning stimulation (HFS) in ACC slices of control mice and mice with peripheral inflammation induced by the injection of complete Freund adjuvant (CFA) to the hind paw. Immunoreactivity of glial fibrillary acidic protein in laminae II–III of the ACC in the CFA‐injected mice was higher than in the control mice. Neuronal excitation in ACC slices from the CFA‐injected mice was gradually increased by HFS, and the magnitude of this long‐term facilitation was greater than in the control mice. The long‐term facilitation in the CFA‐injected mice was inhibited by the astroglial toxin, the N‐methyl‐d‐aspartate (NMDA) receptor antagonist and NMDA receptor glycine binding site antagonist. The increase of intracellular Ca2+ concentration in astrocytes during HFS was higher in the CFA‐injected mice than in the control mice and was inhibited by l‐&agr;‐aminoadipate (l‐&agr;‐AA). These results suggest that the activation of astrocytes in the ACC plays a crucial role in the development of negative emotions and LTP during pain hypersensitivity after peripheral inflammation.

Contribution of anterior cingulate cortex and descending pain inhibitory system to analgesic effect of lemon odor in mice

BackgroundAffections are thought to regulate pain perception through the descending pain inhibitory system in the central nervous system. In this study, we examined in mice the affective change by inhalation of the lemon oil, which is well used for aromatherapy, and the effect of lemon odor on pain sensation. We also examined the anterior cingulate cortex (ACC) and descending pain inhibitory system to such regulation of pain.ResultsIn the elevated plus maze, the time spent in the open arms was increased by inhalation of lemon oil. The pain behavior induced by injection of formalin into the hind paw was decreased. By inhalation of lemon oil, the number of c-Fos expression by formalin injection was significantly increased in the ACC, periaqueductal grey (PAG), nucleu raphe magnus (NRM) and locus ceruleus, and decreased in the spinal dorsal horn (SDH). The destruction of the ACC with ibotenic acid led to prevent the decrease of formalin-evoked nocifensive behavior in mice exposed to lemon oil. In these mice, the change of formalin-induced c-Fos expression in the ACC, lateral PAG, NRM and SDH by lemon odor was also prevented. Antagonize of dopamine D1 receptor in the ACC prevented to the analgesic effect of lemon oil.ConclusionsThese results suggest that the analgesic effect of lemon oil is induced by dopamine-related activation of ACC and the descending pain inhibitory system.

Classification of Skeletal Wireframe Representation of Hand Gesture Using Complex-Valued Neural Network

Complex-valued neural networks (CVNNs), that allow processing complex-valued data directly, have been applied to a number of practical applications, especially in signal and image processing. In this paper, we apply CVNN as a classification algorithm for the skeletal wireframe data that are generated from hand gestures. A CVNN having one hidden layer that maps complex-valued input to real-valued output was used, a training algorithm based on Levenberg Marquardt algorithm (CLMA) was derived, and a task to recognize 26 different gestures that represent English alphabet was given. The initial image processing part consists of three modules: real-time hand tracking, hand-skeletal construction, and hand gesture recognition. We have achieved; (1) efficient and accurate gesture extraction and representation in complex domain, (2) training of the CVNN utilising CLMA, and (3) providing a proof of the superiority of the aforementioned methods by utilising complex-valued learning vector quantization. A comparison with real-valued neural network shows that a CVNN with CLMA provides higher recognition performance, accompanied by significantly faster training. Moreover, a comparison of six different activation functions was performed and their utility is argued.

Layered Ensemble Architecture for Time Series Forecasting

Time series forecasting (TSF) has been widely used in many application areas such as science, engineering, and finance. The phenomena generating time series are usually unknown and information available for forecasting is only limited to the past values of the series. It is, therefore, necessary to use an appropriate number of past values, termed lag, for forecasting. This paper proposes a layered ensemble architecture (LEA) for TSF problems. Our LEA consists of two layers, each of which uses an ensemble of multilayer perceptron (MLP) networks. While the first ensemble layer tries to find an appropriate lag, the second ensemble layer employs the obtained lag for forecasting. Unlike most previous work on TSF, the proposed architecture considers both accuracy and diversity of the individual networks in constructing an ensemble. LEA trains different networks in the ensemble by using different training sets with an aim of maintaining diversity among the networks. However, it uses the appropriate lag and combines the best trained networks to construct the ensemble. This indicates LEAs emphasis on accuracy of the networks. The proposed architecture has been tested extensively on time series data of neural network (NN)3 and NN5 competitions. It has also been tested on several standard benchmark time series data. In terms of forecasting accuracy, our experimental results have revealed clearly that LEA is better than other ensemble and nonensemble methods.

Determining auditory-evoked activities from multiple cells in layer 1 of the dorsal cortex of the inferior colliculus of mice by in vivo calcium imaging

Layer 1 of the dorsal cortex of the inferior colliculus (DCIC) is distinguished from other layers by its cytoarchitecture and fiber connections. However, the information of the sound types represented in layer 1 of the DCIC remains unclear because placing electrodes on such thin structures is challenging. In this study, we utilized in vivo calcium imaging to assess auditory-evoked activities in multiple cells in layer 1 of DCIC and to characterize sound stimuli producing strong activity. Most cells examined showed strong responses to broad-band noise and low-frequency tone bursts of high sound intensity. In some cases, we successfully obtained frequency response areas, which are receptive fields to tone frequencies and intensities, and ~30% of these showed V-shape tunings. This is the first systematic study to record auditory responses of cells in layer 1 of DCIC. These results indicate that cells in this area are selective to tones with low frequency, implying the importance of such auditory information in the neural circuitry of layer 1 of DCIC.

Filamin A-interacting protein (FILIP) is a region-specific modulator of myosin 2b and controls spine morphology and NMDA receptor accumulation

Learning and memory depend on morphological and functional changes to neural spines. Non-muscle myosin 2b regulates actin dynamics downstream of long-term potentiation induction. However, the mechanism by which myosin 2b is regulated in the spine has not been fully elucidated. Here, we show that filamin A-interacting protein (FILIP) is involved in the control of neural spine morphology and is limitedly expressed in the brain. FILIP bound near the ATPase domain of non-muscle myosin heavy chain IIb, an essential component of myosin 2b, and modified the function of myosin 2b by interfering with its actin-binding activity. In addition, FILIP altered the subcellular distribution of myosin 2b in spines. Moreover, subunits of the NMDA receptor were differently distributed in FILIP-expressing neurons, and excitation propagation was altered in FILIP-knockout mice. These results indicate that FILIP is a novel, region-specific modulator of myosin 2b.

Pattern Generation through Feature Values Modification and Decision Tree Ensemble Construction

An ensemble method produces diverse classifiers and combines their decisions for ensemble’s decision. A number nof methods have been investigated for constructing ensemble in which some of them train classifiers with the generated npatterns. This study investigates a new technique of training pattern generation that is easy and effective for ensemble construction. The method modifies feature values of some patterns with the values of other patterns to generate different patterns for different classifiers. The ensemble of decision trees based on the proposed technique was evaluated using a suite of 30 benchmark classification problems, and was found to achieve performance better than or competitive with related conventional methods. Furthermore, two different hybrid ensemble methods have been investigated incorporating the proposed technique of pattern generation with two popular ensemble methods bagging and random subspace method (RSM). It is found that the performance of bagging and RSM algorithms can be improved by incorporating feature values modification with their training processes. Experimental investigation of different types of modification techniques finds that feature values modification with pattern values in the same class is better for generalization.