Computational modeling of reinforcement learning using probabilistic selection task and instructional probabilistic selection task

IntroductionHumans learn how to behave both through rules and instructions as well as through environmental experiences. It has been shown that instructions can powerfully control people's choices, often leading to a confirmation bias.AimTo compare learning parameters in reinforcement learning task with and without instructions.MethodsWe recruited 52 healthy adult control subjects (21 males, 31 females, age 30 ± 6.5 years). Participants completed Repeatable Battery of Neuropsychological Status (RBANSS). Twenty-seven participants completed additionally Probabilistic Selection Task (PST) while twenty-five participants completed Instructional Probabilistic Selection Task (IPST). To analyze learning parameters, we used Q-learning model with 3 parameters: learning rate due to positive and negative reinforcements as well as exploration-exploitation parameter.ResultsBoth groups did not differ with respect to cognitive functioning measured with RBANSS (immediate and delayed memory, visuospatial abilities, language and attention); however, participants who completed PST had trend-level statistically faster learning rates due to positive (P = 0.099) and negative reinforcements (0.057) in comparison to participants who completed IPST. Both groups did not differ with respect to exploration-exploitation parameter (0.409).ConclusionIn healthy adults, interference of confirmation bias can influence learning speed independent of cognitive functioning (immediate and delayed memory, visuospatial abilities, language and attention).Disclosure of interestThe authors have not supplied their declaration of competing interest.

Degradation of Protective Al - Si Coatings during Exploitation of Gas Turbine Blades

Diffusion Al-Si coatings are often used to protect rotor blades of aircraft engines against high-temperature corrosion in environments containing sulfur compounds. Besides other microstructural changes, the degradation of AlSi layers can be indicated by an increasing amount of surface oxide phases and changes in parameters of the layer geometry. In practice, the timetemperature area beyond a critical temperature of the outgoing gas is used as an empirical exploitation parameter D indicating a degradation level. The efficiency of such approach was investigated by analyzing degradation features in the surface layers of rotor blades after exploitations corresponding to different values of D. Determined simple relationships between the relative thickness of degraded layer and the parameter D verify the methodology and yield its clear geometrical interpretation. However, this method fails to provide reasonable information in case when the gas is burning outside the combustion chamber due to a sudden decrease of turbine revolutions.