A Dataset Splitting Based Neural Network Ensemble Method for Tumor Classification

Huijuan Lu; Wutao Chen; Xiaoping Ma; Lu Yi
March 2012
International Journal of Digital Content Technology & its Applic;Mar2012, Vol. 6 Issue 5, p167
Academic Journal
In this paper, we present a neural network ensemble method called DS-NNE that is based on dataset splitting. To get better classification result, the DS-NNE method performs the following tasks: firstly, it performs gene selection with t-test and f-test to remove redundancy genes; secondly, it divides the original training dataset into k disjoint subsets; thirdly, it performs random re-sampling on k-1 out of k subsets to get a training dataset and train a neural network classifier on it, and then repeats the training procedure n times to obtain n neural networks; finally, it predicts the class label of the unknown data with the ensemble classifier through majority vote method. The DS-NNE method has two salient advantages: it is more stable than the bagging and boosting methods; it can obtain higher prediction accuracy compared to other methods. Experiments on five benchmark tumor datasets confirm the validity of the proposed method.


Related Articles

  • Simulation of Significant Wave Height by Neural Networks and Its Application to Extreme Wave Analysis. Aminzadeh-Gohari, A.; Bahai, H.; Bazargan, H. // Journal of Atmospheric & Oceanic Technology;Apr2009, Vol. 26 Issue 4, p778 

    The derivation of the long-term statistical distribution of significant wave heights (Hss) is discussed in this paper. The distribution parameters are estimated using artificial neural networks (ANNs) trained with the help of a simulated annealing algorithm and operated in an autoregressive...

  • Spike-timing-dependent plasticity in a recurrently connected neuronal network with spontaneous oscillations. Gilson, Matthieu; Burkitt, Anthony N.; Grayden, David B.; Thomas, Doreen A.; Van Hemmen, J. Leo // BMC Neuroscience;2009 Supplement 1, Vol. 10, Special section p1 

    Introduction Spike-timing-dependent plasticity (STDP) is believed to generate structure in neuronal networks by modifying synaptic weights depending on the spike-time information contained in the spike trains at the scale of milliseconds. In addition to spiking-rate-based information, STDP...

  • How Gibbs distributions may naturally arise from synaptic adaptation mechanisms. Vasquez, Juan C.; Cessac, Bruno; Rostro-Gonzalez, Horacio; Vieville, Thierry // BMC Neuroscience;2009 Supplement 1, Vol. 10, Special section p1 

    It is assumed that complex perceptual or sensori-motor tasks are the result of neural network dynamics and are expressed by spike trains containing the neural code. Hence, in this context two main questions are (i) How to characterize the statistical properties of sequences the spikes trains...

  • The effects of various spatial distributions of weak noise on rhythmic spiking. Tuckwell, Henry; Jost, Jürgen // Journal of Computational Neuroscience;Apr2011, Vol. 30 Issue 2, p361 

    We consider the response of the classical Hodgkin-Huxley (HH) spatial system in the weak to intermediate noise regime near the bifurcation to repetitive spiking. The deterministic component of the input (signal) is restricted to a small segment near the origin whereas noise, with parameter σ,...

  • Dynamic Finite Size Effects in Spiking Neural Networks. Buice, Michael A.; Chow, Carson C. // PLoS Computational Biology;Jan2013, Vol. 9 Issue 1, p1 

    We investigate the dynamics of a deterministic finite-sized network of synaptically coupled spiking neurons and present a formalism for computing the network statistics in a perturbative expansion. The small parameter for the expansion is the inverse number of neurons in the network. The network...

  • Private versus Public Organizations and High versus Low Experienced Workers' Perceptions of Privatization in Nigeria. Nwankwo, Okechukwu Dominic; Akam, Uche G. // Journal of Emerging Trends in Economics & Management Sciences;Oct2011, Vol. 2 Issue 5, p402 

    This study investigated the perceptions different categories of workers had about the privatization of public organizations in Nigerian. The participants used include the private and public organizations as well as the high experienced and low experienced workers. The workers' perceptions were...

  • Digit Ratio Effects Between Expertise Levels in American Football Players. Schorer, Jörg; Rienhoff, Rebecca; Westphal, Hilmar; Baker, Joseph // Talent Development & Excellence;Oct2013, Vol. 5 Issue 2, p113 

    Previous research highlights the association between 2D:4D and performance in a host of elite sports. This study compared several variables related to the 2D:4D digit ratio among American football players of different skill levels. Onefactorial analyses of variance between four groups were used....

  • Large-Scale Computational Modeling of Genetic Regulatory Networks. M. Stetter; G. Deco; M. Dejori // Artificial Intelligence Review;Oct2003, Vol. 20 Issue 1/2, p75 

    The perhaps most important signaling network in living cells is constituted by the interactions of proteins with the genome – the regulatory genetic network of the cell. From a system-level point of view, the various interactions and control loops, which form a genetic network, represent the...

  • The Prethalamus Is Established during Gastrulation and Influences Diencephalic Regionalization. Staudt, Nicole; Houart, Corinne // PLoS Biology;Apr2007, Vol. 5 Issue 4, p0878 

    The vertebrate neural plate contains distinct domains of gene expression, prefiguring the future brain areas. In this study, we draw an extended expression map of the rostral neural plate that reveals discrete domains inside the presumptive posterior forebrain. We show, by fate mapping, that...


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics