Confidence guides priority between forthcoming tasks

Humans can estimate confidence in their decisions, and there is increasing interest on how this feeling of confidence regulates future behavior. Here, we investigate whether confidence in a perceptual task affects prioritizing future trials of that task, independently of task performance. To do so, we experimentally dissociated confidence from performance. Participants judged whether an array of differently colored circles was closer to blue or red, and we manipulated the mean and variability of the circles’ colors across the array. We first familiarized participants with a low mean low variability condition and a high mean high variability condition, which were matched in performance despite participants being more confident in the former. Then we made participants decide in which order to complete forthcoming trials for both conditions. Crucially, prioritizing one condition was associated with being more confident in that condition compared to the other. This relationship was observed both across participants, by correlating inter-individual heterogeneity in prioritization and in confidence, and within participants, by assessing how changes in confidence with accuracy, condition and response times could predict prioritization choices. Our results suggest that confidence, above and beyond performance, guides prioritization between forthcoming tasks, strengthening the evidence for its role in regulating behavior.

Reconciling Rewards with Predictive State Representations

Predictive state representations (PSRs) are models of controlled non-Markov observation sequences which exhibit the same generative process governing POMDP observations without relying on an underlying latent state. In that respect, a PSR is indistinguishable from the corresponding POMDP. However, PSRs notoriously ignore the notion of rewards, which undermines the general utility of PSR models for control, planning, or reinforcement learning. Therefore, we describe a sufficient and necessary accuracy condition which determines whether a PSR is able to accurately model POMDP rewards, we show that rewards can be approximated even when the accuracy condition is not satisfied, and we find that a non-trivial number of POMDPs taken from a well-known third-party repository do not satisfy the accuracy condition. We propose reward-predictive state representations (R-PSRs), a generalization of PSRs which accurately models both observations and rewards, and develop value iteration for R-PSRs. We show that there is a mismatch between optimal POMDP policies and the optimal PSR policies derived from approximate rewards. On the other hand, optimal R-PSR policies perfectly match optimal POMDP policies, reconfirming R-PSRs as accurate state-less generative models of observations and rewards.

0125 Sleep Deprivation Impairs the Ability to Overcome Pre-Existing Framing Bias

Introduction When presented with a choice between sure gains or losses versus gambles, people tend to select sure gains over gambles, but gambles over sure losses. This pre-existing framing bias is embedded in the Framed Gambling Task (FGT), in which subjects choose between a sure option (gain or loss) and a gamble (card from one of two decks). For optimal performance, subjects need to recognize that one deck (‘good deck’) results in better average outcomes than the other deck (‘bad deck’) and select the gamble or sure option depending on the deck (good/bad) rather than the frame (sure loss/gain). A speeded version of the FGT, with 2s response deadlines to induce time pressure, was used in a laboratory total sleep deprivation (TSD) study to determine the impact of sleep loss on the ability to overcome pre-existing framing bias. Methods Eight-six subjects (ages 21–38; 47 females) were randomized (2:1 ratio) to a TSD condition (n=56) or control condition (n=30). They completed the speeded FGT at 11:00 on the baseline day (session 1), and again the following day (session 2) after 27h of wakefulness (TSD group) or 3h of wakefulness (control group). Performance accuracy was defined in terms of optimal task performance, i.e., gambling when the good deck was presented and not gambling when the bad deck was presented. Each test bout had 72 trials across three trial blocks. Results Accuracy improved across trial blocks (F1,84=20.44, p<0.001). In session 2, the TSD group showed lower accuracy (condition by session interaction: F2,84=4.15, p=0.045) and less improvement across trial blocks (condition by session by trial block interaction: F2,168=3.97, p=0.021) than the control group. Even under TSD, the frequency of response timeouts (RT>2s) was low (<3.5% of trials). Conclusion Sleep deprivation degraded FGT performance under time pressure, indicating reduced ability to overcome pre-existing framing bias. Support PRMRP award W81XWH-16-1-0319

GRAPHICAL PART OF LAND AND BUILDINGS REGISTRY BASED ON AERIAL PHOTOS FROM THE BOARD OF UNMANNED AERIAL VEHICLE (UAV)

The paper attempts to develop a new way of verifying and updating data collected in Land and Property Databases, containing information on land and buildings. The report examines currently existing law regulating for the collection of registration of data, mainly in their geometrical aspect, proposes possible ways of validating these data and enriches with some new elements based on UAS technology. By supplementing the databases with new Land and Property objects the study was prepared, taking into account some new legislative provisions related to the principles and scope of the collected data in the Land and Property's resources. The basic problem with the use of photogrammetry from the UAV level for measuring the location and shape of considered objects is ensuring the proper accuracy. The compliance of accuracy condition and the visibility of the objects makes it possible to significantly supplement the registration data databases with some new elements such as terraces, verandas, stairs, etc. The paper discusses these issues and presents the results performed on real objects, together with their accuracy rating. It has been found that images made from low altitude can be used to measure new object classes, update land and buildings database, and also, to a limited extent, validate Land and Property Databases for another, from the up-visible objects.

Study on Process Parameters in Fiber Laser Micro-Cutting Technology

Sections: Fiber laser is a widely used “tool” for micro-machining various materials, in high accuracy condition. Still, there are not set rules and / or principles, so that to be sure of obtaining the required geometric precision for the part. This paper presents some attempts for setting procedures and parameters values domains, so that to further develop technologies for fiber laser micro-drilling of stainless steel sheets, with thickness of 0.5 mm.

Visual and Acoustic Processing in Making Same-Different Judgments

The effect of speed/accuracy instructions on preference for coding format was studied in a sequential same-different task. Visual coding was present in both the speed- and the accuracy-instruction conditions, but acoustic coding was present only in the speed condition, and only when that condition occurred before the accuracy condition. Results are discussed in relation to the minimal-processing principle and the temporal development view of coding preference.