Date: Mon, 04 Apr 2005 21:34:12 +0200
Dear Prof. Heinz,
I regret to inform you that your paper: * "Interference Networks to Model Nerve System in Structure and Behaviour" has been rejected for IWINAC2005 conference. Please, note that the rejection decision has been based on the recommendations of the referees who have used several criteria such as technical quality, fitness to IWINAC2005 topics, newness, clarity, format requirements, etc.
The opinions of the referees have also been weighted according to the degree of expertise declared about themselves.
In some cases, the referees have added a few specific comments on your paper. All these comments are now included together here for your information. Please, note that in some cases the referees suggest a set of improvements, but unfortunately, changes on the paper are not considered after the evaluation.
I value your effort and interest in the IWINAC conference and encourage you to apply again in the next venue, if this is of your interest.
IWINAC2005 General Chairman
Comments for authors about paper (ref. 133)
"Interference Networks to Model Nerve System in Structure and Behaviour":
- The paper introduces a new kind of neural net, called "Inference network", aimed to model the nervous system and to be applied in technological problems. These networks are based on the concept of wave interference, and considers the neural signals as electromagnetic waves [deep misunderstanding: interference nets specify a generalized, abstract field theory basing ONLY on space-time functions - without any materialistic background. You can use it for nerve nets, for acoustic or electric cameras, for locus calculations, for sound maps, map- (image-) projections, feedback control loops or for convolution: the interference of a timefunction-wave with a negative argument timefunction-wave], and post-synaptic sites as "crossroads", where the mentioned waves "crash" with each other, thus producing an interference phenomenon. The metaphor is used to derived different types of spatio-temporal coding [Yes, we speak about spatio-temporal coding in abstract, generalized form - but we speak about more:
about code-selective, space-selective and frequency-selective systems, about control loop theory and about a generalized wave theory combined in a simple approach that is called interference network].
1. Physiological plausibility of the model. There are too many fundamental flaws in the proposed model: 1. Spikes cannot be considered electromagnetic waves [see above]. The last ones are defined as propogation of energy through the air, while spikes are generated in a completely different medium and travels following the neural axon [It is not the question of carrier material, it is only the question of velocity of information flow or timefunction-wave flow] Moreover, there is no such concept as wavelength in a train of spikes because this type of event is not an oscillatory phenomena [Pulse interference has nothing to do with oscillation - example: crash or match of two cars on a crossroad can be calculated using a simplest interference net - if we know positions, velocities, times and wavelengths (length of the cars)].
2. The postsinaptic dendrites are not only crossroads to which axons converge. We know that the integration of postsynaptic potentials is a complex phenomena that consists on a spatiotemporal integration of postsynaptic potentials. Parameters such as synaptic constant times, activation thresholds, synaptic reverse potentials, amount of released neurotransmitter, etc., determine the result of such integration. The interference of electromagnetic waves does not represent the reality of what happens in a real dendrite [See above. The abstraction of fast pulse, fast car, fast animal, fast walk is velocity = ds/dt. The abstraction of a complexest moving, irregular, inhomogeneous system or net is interference network].
3. Lack of physiological evidences. The model should be supported by experimental evidences. Nothing is found throughout the paper
[I write about pulses, wavelength, distances and velocities in nerve system - sorry, no evidence?].
4. No comparison with similar work. There are abundant literature about the problem of temporal coding, synchronization, integrate and fire processes, etc. There is no discussion in which the proposed model is compared with the existing approaches.
[How to compare any abstraction (-> velocity, interference integral...) with a specific behaviour of some neurons? It is a different philosophical level]
5. Paper style and format. The quality of the paper is meager: 1. It does not follow the conference format, 2. the english is poorly written and shows a myriad of mistakes
[thats a real problem: maybe I should write following papers in german] and typos, 3. the written style is informal, to say the least. Overall, the appearance is more of the type of personal notes than of an scientific paper.
- +++ Unsolved format or file problems: Out of format; Latex Sources and figures not found