Abstract
This chapter investigates the need for designers and users to be able to customize CogInfoCom channels. It is argued that the availability of tools for this purpose is important due to the specificities of the CogInfoCom modality that is used—especially with respect to the input device and the noise level characteristic of the transfer medium. However, the task of creating such a model is rendered difficult due to the fact that the function which links all possible combinations of generation parameter values to perceptual qualities (referred to as f eval in Chap. 9) is both difficult to compute and also practically impossible to invert. One possible solution to this challenge is to apply a tuning model that allows users to interactively explore the parametric space used to generate CogInfoCom messages. The chapter introduces the spiral discovery method (SDM)—a tuning model that fulfills these requirements and also empirically aims to support flexibility and interpretability.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Since the f eval function introduced in Chap. 9 is not analytically available and also not invertible, tuning will by necessity resemble a searching process rather than a computation of parameter values from a quantification of a desired perceptual quality.
- 2.
From an engineering perspective, this point can be appreciated if one considers how difficult it used to be to manually tune a radar locator. Although today this is a task that can be performed through automation, a few decades ago it was a task for humans to tune the elevation, azimuth and carrier frequency of the radar locator. This required the use of just three controls; nevertheless, it was a hugely difficult task that demanded much practice and attention.
- 3.
It is also important to note that in general it is difficult to specify in the first place what perceptually optimal CogInfoCom messages would be like.
- 4.
From the user’s perspective, the direction of the hyper-spiral is a transparent parameter, as it will coincide with the direction of the principal component of a set of “control” points.
- 5.
Note that the interactive scheme described here is somewhat relevant to the paradigm of Interactive Evolutionary Computation, as described by Takagi (2001), in the sense that the user’s subjective evaluations are used to guide an iterative search process.
- 6.
The “velocity” parameter of SDM is set transparently to the user, as it influences the sensitivity of the “distance” parameter, but does not influence the direction of parameter discovery.
References
Baranyi P (2004) TP model transformation as a way to LMI based controller design. IEEE Trans Ind Electron 51(2):387–400
Baranyi P (2014) The Generalized TP model transformation for TS fuzzy model manipulation and generalized stability verification. IEEE Trans Fuzzy Syst
Baranyi P, Yam Y, Varlaki P (2006) Tensor product model transformation in polytopic model-based control. CRC/Taylor & Francis, Boca Raton/London
Csapo A, Baranyi P (2012c) The spiral discovery method: an interpretable tuning model for CogInfoCom channels. J Adv Comput Intell Intell Inform 16(2):358–367
Ishteva M, Lathauwer L, Absil P, van Huffel S (2008) Dimensionality reduction for higher-order tensors: algorithms and applications. Int J Pure Appl Math 42(3):337–343
de Lathauwer L, Moor BD, Vandewalle J (2000) A multi linear singular value decomposition. SIAM J Matrix Anal Appl 21(4):1253–1278
Takagi H (2001) Interactive evolutionary computation: fusion of the capabilities of EC optimization and human evaluation. Proc IEEE 89(9):1275–1296
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Baranyi, P., Csapo, A., Sallai, G. (2015). Tunability of CogInfoCom Channels. In: Cognitive Infocommunications (CogInfoCom). Springer, Cham. https://doi.org/10.1007/978-3-319-19608-4_10
Download citation
DOI: https://doi.org/10.1007/978-3-319-19608-4_10
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-19607-7
Online ISBN: 978-3-319-19608-4
eBook Packages: EngineeringEngineering (R0)