1 |
An analytical framework for learning systemsHolte, R. C. January 1988 (has links)
No description available.
|
2 |
A model of information growthObeid, N. January 1987 (has links)
No description available.
|
3 |
Informačně-teoretické vlastnosti vybraných stochastických neuronálních modelů / Information-theoretic properties of selected stochastic neuronal modelsBárta, Tomáš January 2018 (has links)
According to the classical efficient-coding hypothesis, biological neurons are naturally adapted to transmit and process information about the stimulus in an optimal way. Shannon's information theory provides methods to compute the fundamental limits on maximal information transfer by a general system. Understanding how these limits differ between different classes of neurons may help us to better understand how sensory and other information is processed in the brain. In this work we provide a brief review of information theory and its use in computational neuroscience. We use mathematical models of neuronal cells with stochastic input that realistically reproduce different activity patterns observed in real cortical neurons. By employing the neuronal input-output pro- perties we calculate several key information-theoretic characteristics, including the information capacity. In order to determine the information capacity we propose an iterative extension of the Blahut-Arimoto algorithm that generalizes to continuous input channels subjected to constraints. Finally, we compare the information optimality conditions among different models and parameter sets. 1
|
4 |
Diffusion Mediated Signaling: Information Capacity and Coarse Grained RepresentationsGarvey, Matthew Thomas 02 February 2009 (has links)
No description available.
|
5 |
Žmogaus sensomotorinių sistemų informacinis vertinimas / Information evaluation of the human sensorimotor systemsŠimaitis, Andrius 28 August 2009 (has links)
Šiame tyrinėjime informacijos teorijos sąvokos yra panaudotos nustatyti žmogaus diskretinių ir tolydinių akių ir rankos sekamųjų judesių kontrolės sistemas. Šiame tyrime taikinio šokinėjimas iš vienos padėties į kitą (diskretaus sekimo eksperimentas) arba tolydi taikinio judėjimo trajektorija (tolydaus sekimo eksperimentas) ekrane yra apibrėžta kaip įėjimo ar šaltinio informacija, akies ar rankos sekimo trajektorija ekrane - kaip išvesties informacija ir skirtumas tarp jų - kaip prarasta informacija. Informacijos kiekis perduotas per akių ar rankos sensomotorinius kanalus apibrėžiamas kaip skirtumas tarp įvesties ir prarastos informacijos kiekių. Akių ar rankos sensomotorinių sistemų kanalo informacijos geba yra apibrėžiama kaip maksimali perduota informacija per sensomotorinį kanalą per laiko vienetą. Eksperimentinis tyrimas leidžia formuluoti, kad didžiausia diskretaus sekamųjų akių judesių informacijos praleidžiamoji geba yra 4,9 bitai/sek, kai tarpsakadinis intervalas yra 0,3 sekundės. Diskretaus rankos sekimo didžiausia sensomotorinio kanalo informacijos geba yra 6,5 bitai/sek, kai tarpsakadinis intervalas yra 0,25 sekundės. Kai taikinys juda tolydžiai, žmogaus akių ir rankos sensomotorinė kontrolės sistema turi didžiausią kanalo gebą 7,9 bitai/sek ir 6,5 bitai/sek atitinkamai, kai taikinio didžiausias greitis 30 laipsn/sek. / In this research information theory concepts are used to determinate discreet and continuous eye and hand of the human movements control systems. In this research target jumping from one position to another (discreet tracking experiment) or slow target movement trajectory (continuous tracking experiment) on the screen is defined as input or source information, eye or hand response trajectory on the screen – as output information and difference between them – as lost information. Information amount transferred over eye or hand sensorimotor channels us defined as difference between input and lost information rates. Channel information capacity of the eye or hand sensorimotor system is defined as maximum transferred information over sensorimotor channel during unit of time. Experimental investigation let us formulate that largest channel information capacity of the discreet tracking eye movements is 4,9 bits/sec, when intersaccadic interval is 0,3 sec. Largest channel information capacity of the discreet tracking hand is 6,5 bits/sec, when intersaccadic interval is 0,25 sec. During human’s eye and hand tracking responses to the continuously moving target, eye and hand sensorimotor control systems have largest channel capacities 7,9 bits/sec and 6,5 bits/sec respectively, when target movement largest velocity 30 deg/sec.
|
6 |
A Ray-Based Investigation of the Statistical Characteristics and Efficient Representation of Multi-Antenna Communication ChannelsGerman, Gus Ryan 12 July 2004 (has links) (PDF)
Multi-antenna communication systems are attracting research interest as a means to increase the information capacity, reliability, and spectral efficiency of wireless information transfer. Ray-tracing methods predict the behavior of wireless channels using a model of the propagation environment and are a low-cost alternative to direct measurements. We use ray tracing simulations to validate the statistical time and angle of arrival characteristics of an indoor multipath channel and compare model parameter estimates with estimates derived from channel sounding measurements. Ray tracing predicts the time and angle clustering of multipaths observed in the measurements and provides model parameter estimates which are closely correlated with measured estimates. The ray tracing parameters relating to power characteristics show more deviation from measurements than the time and angle related parameters. Our results also indicate that the description of reflective scatterers in the propagation environment is more important to the quality of the predicted statistical behavior than the description of bulk materials. We use a ray synthesis model to investigate means of efficiently representing the channel for feedback information to the transmitter as a means to increase the information capacity. Several methods of selecting the ray-model feedback information are demonstrated with results from simulated and measured channels. These results indicate that an ESPRIT algorithm coupled with ad hoc transmit/receive pairing can yield better than 90% of the ideal waterfilling capacity when adequate training-based channel estimates are available. Additionally, we investigate a covariance feedback method for providing channel feedback for increased capacity. Both the ray-based and covariance-based feedback methods yield their highest capacity improvements when the signal to noise ratio is low. This results because of the larger benefit of focusing transmit power into the most advantageous eigenmodes of the channel when fewer eigenmodes have power allocated to them by the waterfilling capacity solution. In higher signal to noise ratio cases, more eigenmodes of the channel receive power when waterfilling, and the capacity improvement from feedback information decreases relative to a uniform power allocation. In general, ray model feedback methods are preferable because the covariance feedback quickly requires higher computational effort as the array sizes increase and typically results in lower capacity for a given amount of feedback information.
|
Page generated in 0.081 seconds