Research output: Other contribution

**Discretisation of Data in a Binary Neural k-Nearest Neighbour Algorithm.** / Hodge, Victoria Jane; Austin, Jim.

Research output: Other contribution

Hodge, VJ & Austin, J 2012, *Discretisation of Data in a Binary Neural k-Nearest Neighbour Algorithm*. Department of Computer Science, University of York, UK.

Hodge, V. J., & Austin, J. (2012, Jun 1). Discretisation of Data in a Binary Neural k-Nearest Neighbour Algorithm.

Hodge VJ, Austin J. Discretisation of Data in a Binary Neural k-Nearest Neighbour Algorithm. 2012.

@misc{b9cefe73a557406d9c852c02e045e63a,

title = "Discretisation of Data in a Binary Neural k-Nearest Neighbour Algorithm",

abstract = "This paper evaluates several methods of discretisation (binning) within a k-Nearest Neighbour predictor. Our k-NN is constructed using binary neural networks which require continuous-valued data to be discretised to allow it to be mapped to the binary neural framework. Our approach uses discretisation coupled with robust encoding to map data sets onto the binary neural network. In this paper, we compare seven unsupervised discretisation methods for retrieval accuracy (prediction accuracy) across a range of well-known prediction data sets comprising time-series data. We analyse whether there is an optimal discretisation configuration for our k-NN. The analyses demonstrate that the configuration is data specific. Hence, we recommend running evaluations of a number of configurations, varying both the discretisation methods and the number of discretisation bins, using a test data set. This evaluation will pinpoint the optimum configuration for new data sets.",

keywords = "k-Nearest Neighbour, binary neural network, discretisation, binning, quantisation",

author = "Hodge, {Victoria Jane} and Jim Austin",

year = "2012",

month = jun,

day = "1",

language = "English",

volume = "Technical Report YCS-2012-473",

type = "Other",

}

TY - GEN

T1 - Discretisation of Data in a Binary Neural k-Nearest Neighbour Algorithm

AU - Hodge, Victoria Jane

AU - Austin, Jim

PY - 2012/6/1

Y1 - 2012/6/1

N2 - This paper evaluates several methods of discretisation (binning) within a k-Nearest Neighbour predictor. Our k-NN is constructed using binary neural networks which require continuous-valued data to be discretised to allow it to be mapped to the binary neural framework. Our approach uses discretisation coupled with robust encoding to map data sets onto the binary neural network. In this paper, we compare seven unsupervised discretisation methods for retrieval accuracy (prediction accuracy) across a range of well-known prediction data sets comprising time-series data. We analyse whether there is an optimal discretisation configuration for our k-NN. The analyses demonstrate that the configuration is data specific. Hence, we recommend running evaluations of a number of configurations, varying both the discretisation methods and the number of discretisation bins, using a test data set. This evaluation will pinpoint the optimum configuration for new data sets.

AB - This paper evaluates several methods of discretisation (binning) within a k-Nearest Neighbour predictor. Our k-NN is constructed using binary neural networks which require continuous-valued data to be discretised to allow it to be mapped to the binary neural framework. Our approach uses discretisation coupled with robust encoding to map data sets onto the binary neural network. In this paper, we compare seven unsupervised discretisation methods for retrieval accuracy (prediction accuracy) across a range of well-known prediction data sets comprising time-series data. We analyse whether there is an optimal discretisation configuration for our k-NN. The analyses demonstrate that the configuration is data specific. Hence, we recommend running evaluations of a number of configurations, varying both the discretisation methods and the number of discretisation bins, using a test data set. This evaluation will pinpoint the optimum configuration for new data sets.

KW - k-Nearest Neighbour

KW - binary neural network

KW - discretisation

KW - binning

KW - quantisation

M3 - Other contribution

VL - Technical Report YCS-2012-473

CY - Department of Computer Science, University of York, UK

ER -