Effects of feedback delay on learning

In this paper, the effect of feedback delay and channel gain difference (CGD) on the performance of distributed transmit diversity (DTD) systems are analyzed over Rayleigh fading channels. We derive new expressions for the average error rate of DTD in the absence of receiver phase compensation (RPC), which is useful for a large number of modulation schemes. Furthermore, we analyze the effect of feedback delay and CGD on DTD compared to spatially correlated co-located transmit diversity (CTD) systems. Analytical results enable us to verify the fact that in DTD without RPC as CGD between antennas increases the performance degrades. Via extensive performance evaluations for various cases, we quantitatively analyze the sensitivity of the performance gap between CTD and DTD in absence of RPC according to feedback delay. The results show that, in without RPC case DTD is robust to feedback delay compared to correlated CTD. We also present Monte Carlo simulation results as verification of our analytical results.

This paper is intended to model a decision maker as a rational probabilistic decider (RPD) and to investigate its behavior in stationary and symmetric Markov switch environments. RPDs take their decisions based on penalty functions defined by the environment. The quality of decision making depends on a parameter referred to as level of rationality. The dynamic behavior of RPDs is described by an ergodic Markov chain. Two classes of RPDs are considered—local and global. The former take their decisions based on the penalty in the current state while the latter consider all states. It is shown that asymptotically (in time and in the level of rationality) both classes behave quite similarly. However, the second largest eigenvalue of Markov transition matrices for global RPDs is smaller than that for local ones, indicating faster convergence to the optimal state. As an illustration, the behavior of a chief executive officer, modeled as a global RPD, is considered, and it is shown that the company performance may or may not be optimized—depending on the pay structure employed. While the current paper investigates individual RPDs, a companion paper will address collective behavior.

In this letter, we investigate the effect of feedback delay on the performance of a dual-hop amplify-and-forward (AF) relay network with beamforming in the presence of multiple interferers over Rayleigh fading channels. Specifically, we derive closed-form expressions for the outage probability (OP) and the symbol error rate (SER). Furthermore, to render insights into the effect of feedback delay and interference on the network performance, asymptotic OP and SER are also presented. These asymptotic expressions are very tight in the high signal-to-noise ratio regime, readily enabling us to obtain the diversity and coding gains of the considered network.

In this paper, the outage performance of multiuser multiple-input single-output (MISO) systems exploiting joint spatial and multiuser diversities is investigated for Rayleigh fading channels with outdated feedback. First, we derive closed-form exact outage probabilities for the joint diversity schemes that combine user scheduling with different spatial diversity techniques, including: 1) transmit maximum-ratio combining (TMRC); 2) transmit antenna selection (TAS); and 3) orthogonal space-time block coding (OSTBC). Then the asymptotic outage probabilities are analyzed to gain more insights into the effect of feedback delay. It is observed that with outdated feedback, the asymptotic diversity order of the multiuser OSTBC (M-OSTBC) scheme is equal to the number of transmit antennas at the base station, while that of the multiuser TMRC (M-TMRC) and the multiuser TAS (M-TAS) schemes reduce to one. Further by comparing the asymptotic outage probabilities, it is found that the M-TMRC scheme outperforms the M-TAS scheme, and the M-OSTBC scheme can perform best in the outage regime of practical interest when the feedback delay is large. Theoretical analysis is verified by simulation results.

The authors investigate deployment of link adaptation techniques to enhance the performance of real-time video communications over time-varying error-prone environments. A novel link adaptation algorithm is proposed and analysed. Two approaches for link adaptation are investigated. The first approach aims to optimise video quality for a fixed channel allocation. The design criterion of the second approach is to maximise the overall access network throughput while maintaining the required video quality. The effectiveness of the algorithms is demonstrated for MPEG-4 coded video transmissions over simulated EGPRS networks. In addition, the effects of feedback delay, a noisy feedback channel and bursty channel errors on algorithm performance are investigated.

This paper investigates the outage probability of the multiuser transmit antenna selection/maximal-ratio combining (TAS/MRC) system over Rayleigh fading channels with outdated feedback. The exact outage probability is first derived. Then the asymptotic outage probability is analyzed to gain more insights into the effect of feedback delay. The analytical results show that: 1) increasing the number of users has the same effect on the outage probability as increasing the number of transmit antennas at the base station; 2) with outdated feedback, the asymptotic diversity order of the multiuser TAS/MRC system reduces to the number of receive antennas at the users. Theoretical analysis is verified by simulation results.

Allocating wireless resources often relies on the accurate feedback of channel state information (CSI). In this paper, we study how strongly a multiuser OFDM downlink with single-cell scheduling and channel estimation suffers from feedback delay. Unlike previous work, we study this degradation for optimal joint power and rate allocation under fairness constraints. Comparing the performance of the ideal case to delayed CSI shows that (i) adjusting the schedulers fairness cannot mitigate the strong performance loss due to feedback delay but that (ii) simple linear channel prediction is a powerful tool to do so.

In this paper, the effect of feedback delays on the performance of multiple-input multiple-output antenna amplify-and-forward relay networks with the best transmit/receive antenna pair selection over Rayleigh fading is studied. The cumulative distribution function and the moment generating function of the end-to-end signal-to-noise ratio (SNR) are derived. Closed-form expressions for the outage probability, average symbol error rate (SER), and the SNR moments are also derived. To gain further insights, the asymptotic outage probability, average SER, diversity order, and coding gain are presented. Numerical results and Monte-Carlo simulations are provided to illustrate the detrimental effects of feedback delays on the system performance and to verify the accuracy of our analysis.

This research studies the strategic level decision-making in the biofuel industry by considering the effect of investment and disinvestment delay. Under uncertain environments, the delay causes additional risk exposure to decision makers, and the optimal decision reflects the effect of delay. This research provides the concise and comprehensive profile of the optimal investment and disinvestment policy by taking investment and disinvestment delay into consideration. The results show that the trend of fuel price plays a crucial role to determine the effect of delay. If the trend is favorable to the alternative biofuel plant, longer exercise delays provide the decision makers an incentive to act more quickly, and shorter delays provide an opposite incentive. By applying the suggested model to the bioenergy investments cases in the state of Georgia, the United States, we provide managerial insights for the strategic decision makers in the biofuel industry.

Introduction: Training in a certain skill not only culminates in better motor performance but also leads to neural adjustments and regulations in different brain regions. Aim: The present study aimed to psychologically assess event-related motor potential in motor-related cortical regions during simple and complicated tasks. Method: The current research was of a semi-experimental type with a single-phase post-test design without a control group. The participants of the present study consisted of all 20-32-year-old right-handed male table tennis players in Tehran, in 2019. A total of 8 samples were selected via convenience sampling. Data were collected at the National Brain Mapping Laboratory. Annett Hand Preference Questionnaire (1970), Ocular Dominance Testing Card (1910), the Newgy Robo Pong Model 2050 Digital Table Tennis Robot, and the Specific Test for Table Tennis Stroke (2016) were used to record speed from Radar gun, and cameras and judges’ scoring were also used to measure accuracy. Moreover, the two-way analysis of variance (ANOVA) statistical test was employed to assess the differences in two levels of task complexity in SPSS software version 24. Results: The results showed that the motor potential in the primary motor cortex region had a greater range in the simple task than in the complicated task (P=0.012); however, the premotor (P=0.008) and the posterior parietal cortical regions (P=0.026) of the complicated task had larger ranges than the simple task. Conclusion: According to the results of this study, the motor potential range is dependent on the task complexity level. Therefore, it is suggested that tasks with different complexity levels be taken into account in investigating event-dependent potentials.

Reproduction, growth, stress, and metabolism are determined by endocrine/neuroendocrine systems that regulate circulating hormone concentrations. All these systems generate rhythms and changes in hormone pulsatility observed in a variety of pathophysiological states. Thus, the output of endocrine/neuroendocrine systems must be regulated within a narrow window of effective hormone concentrations but must also maintain a capacity for plasticity to respond to changing physiological demands. Remarkably most endocrinologists still have a "textbook" view of endocrine gland organization which has emanated from 20th century histological studies on thin 2D tissue sections. However, 21st -century technological advances, including in-depth 3D imaging of specific cell types have vastly changed our knowledge. We now know that various levels of multicellular organization can be found across different glands, that organizational motifs can vary between species and can be modified to enhance or decrease hormonal release. This article focuses on how the organization of cells regulates hormone output using three endocrine/neuroendocrine glands that present different levels of organization and complexity: the adrenal medulla, with a single neuroendocrine cell type; the anterior pituitary, with multiple intermingled cell types; and the pancreas with multiple intermingled cell types organized into distinct functional units. We give an overview of recent methodologies that allow the study of the different components within endocrine systems, particularly their temporal and spatial relationships. We believe the emerging findings about network organization, and its impact on hormone secretion, are crucial to understanding how homeostatic regulation of endocrine axes is carried out within endocrine organs themselves. © 2022 American Physiological Society. Compr Physiol 12:3371-3415, 2022.

This paper presents and compares four models of games and risk analyses designed to support strategic and policy decisions, three focusing on national security issues and one on project management. They share a common core of probability, linked decisions among the parties involved, and risks to a principal decision maker. Their structure is based on systems and decision analysis. Their level of complexity depends on the strategies, the environment, and assumptions of variation over time of probabilities, preferences, and options. They are part of the field of analytics and some of its real-life applications. The first model is based on a one-move game, in which the United States faces risks of terrorist attacks by several possible groups using various types of weapons. The result is a probabilistic ranking, at a given time, of the threat posed by these weapons. The second model is a dynamic simulation of counterterrorism scenarios in an alternate game between a government and a terrorist group. The objective is to compare the stabilizing effects of different short- and long-term government strategies. The third model is a dynamic evaluation of nuclear counterproliferation strategies involving an analysis of the weapon development program of a particular country with evolving intent and capabilities and of the effectiveness of different U.S. strategies to prevent or delay its success. The fourth model is a principal–agent representation of the development of an engineered system, in which an agent in charge of part of the project may consider meeting a deadline by cutting corners if he falls behind schedule, generally increasing the system failure probability. The goal is to support the decisions of the manager in setting constraints and incentives. This paper shows how a set of similar game and risk analysis models at different levels of complexity can provide valuable insights to decision makers, both in national security and management situations, and help them avoid mistakes such as excessive focus on the short term and underestimation of dependencies. It compares their capabilities, including the number of moves, dynamics of the underlying situation, possible changes of context, actors' preferences and strategic options, and risk characterization.

Should employees be allowed to discuss their pay with other employees? Should managers explain the logic underlying organizational pay structures and decisions to their employees? Should companies disclose the mean or median rates of pay for particular positions, or perhaps even individual employees’ actual pay? Concerns over gender and racial disparities are increasingly driving organizations to liberalize restrictions on pay communication. But the forces driving such moves have, to date, been based more on conjecture than science. Accordingly, the aim of this book is to offer evidence-based insights into how pay communication policies and practices impact outcomes at individual, organizational, and societal levels. For example, at the individual level, the book examines the impact of pay transparency on task performance, helping and collaboration, and counterproductive work behavior. At the organizational level, it examines how pay transparency impacts employee turnover rates, pay dispersion, and reliance on idiosyncratic deals. Finally, from a society level, the book explores pay transparency’s impact on gender pay disparities. Summarizing findings from the recent surge in pay transparency research, this book aims to help employees, managers, and policymakers better understand how pay communication policies and practices may be leveraged to enhance organizational performance and address social inequality, while at the same time highlighting for them why and when such practices can generate potentially harmful, unintended consequences.

Modeling is essential to integrating knowledge of human physiology. Comprehensive self-consistent descriptions expressed in quantitative mathematical form define working hypotheses in testable and reproducible form, and though such models are always "wrong" in the sense of being incomplete or partly incorrect, they provide a means of understanding a system and improving that understanding. Physiological systems, and models of them, encompass different levels of complexity. The lowest levels concern gene signaling and the regulation of transcription and translation, then biophysical and biochemical events at the protein level, and extend through the levels of cells, tissues and organs all the way to descriptions of integrated systems behavior. The highest levels of organization represent the dynamically varying interactions of billions of cells. Models of such systems are necessarily simplified to minimize computation and to emphasize the key factors defining system behavior; different model forms are thus often used to represent a system in different ways. Each simplification of lower level complicated function reduces the range of accurate operability at the higher level model, reducing robustness, the ability to respond correctly to dynamic changes in conditions. When conditions change so that the complexity reduction has resulted in the solution departing from the range of validity, detecting the deviation is critical, and requires special methods to enforce adapting the model formulation to alternative reduced-form modules or decomposing the reduced-form aggregates to the more detailed lower level modules to maintain appropriate behavior. The processes of error recognition, and of mapping between different levels of model complexity and shifting the levels of complexity of models in response to changing conditions, are essential for adaptive modeling and computer simulation of large-scale systems in reasonable time.

Reinforcement learning (RL) has become widely adopted in robot control. Despite many successes, one major persisting problem can be very low data efficiency. One solution is interactive feedback, which has been shown to speed up RL considerably. As a result, there is an abundance of different strategies, which are, however, primarily tested on discrete grid-world and small scale optimal control scenarios. In the literature, there is no consensus about which feedback frequency is optimal or at which time the feedback is most beneficial. To resolve these discrepancies we isolate and quantify the effect of feedback frequency in robotic tasks with continuous state and action spaces. The experiments encompass inverse kinematics learning for robotic manipulator arms of different complexity. We show that seemingly contradictory reported phenomena occur at different complexity levels. Furthermore, our results suggest that no single ideal feedback frequency exists. Rather that feedback frequency should be changed as the agent’s proficiency in the task increases.