-
Distinguishing Learning Rules with Brain Machine Interfaces
Authors:
Jacob P. Portes,
Christian Schmid,
James M. Murray
Abstract:
Despite extensive theoretical work on biologically plausible learning rules, clear evidence about whether and how such rules are implemented in the brain has been difficult to obtain. We consider biologically plausible supervised- and reinforcement-learning rules and ask whether changes in network activity during learning can be used to determine which learning rule is being used. Supervised learn…
▽ More
Despite extensive theoretical work on biologically plausible learning rules, clear evidence about whether and how such rules are implemented in the brain has been difficult to obtain. We consider biologically plausible supervised- and reinforcement-learning rules and ask whether changes in network activity during learning can be used to determine which learning rule is being used. Supervised learning requires a credit-assignment model estimating the map** from neural activity to behavior, and, in a biological organism, this model will inevitably be an imperfect approximation of the ideal map**, leading to a bias in the direction of the weight updates relative to the true gradient. Reinforcement learning, on the other hand, requires no credit-assignment model and tends to make weight updates following the true gradient direction. We derive a metric to distinguish between learning rules by observing changes in the network activity during learning, given that the map** from brain to behavior is known by the experimenter. Because brain-machine interface (BMI) experiments allow for precise knowledge of this map**, we model a cursor-control BMI task using recurrent neural networks, showing that learning rules can be distinguished in simulated experiments using only observations that a neuroscience experimenter would plausibly have access to.
△ Less
Submitted 16 October, 2022; v1 submitted 27 June, 2022;
originally announced June 2022.
-
Glass engineering to enhance Si solar cells: a case study of Pr$^{3+}$-Yb$^{3+}$ codoped tellurite-tungstate as spectral converter
Authors:
Maiara Mitiko Taniguchi,
Vitor Santaella Zanuto,
Pablo Portes,
Luis Carlos Malacarne,
Nelson Guilherme Astrath,
Jorge Diego Marconi,
Marcos Paulo Belançon
Abstract:
Spectral converters are known to increase photovoltaic energy conversion by minimizing losses due to fundamental non-absorption and thermalization processes, and have been suggested to surpass the Shockley-Queisser efficiency limit in single junction solar cells. Here we present a detailed spectroscopic study of photoluminescence in tellurite-tungstate glasses doped and codoped with…
▽ More
Spectral converters are known to increase photovoltaic energy conversion by minimizing losses due to fundamental non-absorption and thermalization processes, and have been suggested to surpass the Shockley-Queisser efficiency limit in single junction solar cells. Here we present a detailed spectroscopic study of photoluminescence in tellurite-tungstate glasses doped and codoped with $Pr^{3+}-Yb^{3+}$ and $Ag$ nanoparticles. The energy transfer mechanisms between $Pr^{3+}$ and $Yb^{3+}$ are discussed based on the near infrared emission under excitation at $442$ nm and on the upconversion emission under excitation at $980$ nm. Fluorescence quenching of $^2 F_{5/2}$ level of $Yb^{3+}$ is observed by increasing the concentration of $Pr^{3+}$, as well as by the addition of $Ag$ nanoparticles. In addition, a discussion on the potential of this glass to increase energy production in spectral converters is presented. The results suggest that the few undesirable energy transfer processes occurring in this material are difficult to be controlled or eliminated properly, resulting in intrinsic losses. This discussion is extended to the potential of glass science to enhance energy production in solar cells, showing that newer designs such as bifacial cells may facilitate the exploration of glasses other than soda-lime for mass production of solar cells. The focus on extending the lifespan by reducing UV induced degradation seems to be a more effective approach than the development of spectral converters for Si solar cells.
△ Less
Submitted 14 April, 2020; v1 submitted 27 September, 2019;
originally announced September 2019.
-
Considerations on Interdisciplinary Instruction and Design Influenced by Adaptive Learning. A Case Study Involving Biology, Computer Science, Mathematics, and Statistics
Authors:
Karen Aguar,
Charles C. Sanchez,
Diego Boada Beltran,
Saeid Safaei,
Mehdi Asefi,
Jonathan Arnold,
Pedro Portes,
Hamid R. Arabnia,
Juan B. Gutierrez
Abstract:
ALICE (Adaptive Learning for Interdisciplinary Collaborative Environments) is an open-source web based adaptive learning system designed for interdisciplinary instruction. ALICE has the potential to transform education by empowering transdisciplinary knowledge acquisition. This is particularly important in fields that accept newcomers with diverse scholastic backgrounds, e.g. Systems Biology. With…
▽ More
ALICE (Adaptive Learning for Interdisciplinary Collaborative Environments) is an open-source web based adaptive learning system designed for interdisciplinary instruction. ALICE has the potential to transform education by empowering transdisciplinary knowledge acquisition. This is particularly important in fields that accept newcomers with diverse scholastic backgrounds, e.g. Systems Biology. With traditional interdisciplinary instruction, the instructor must cover pre-requisite information from multiple disciplines to ensure all students begin at a common baseline - slowing the learning process. With ALICE, students follow a personalized syllabus based on their previous knowledge and work towards individual goals. Implementing an adaptive learning system in an interdisciplinary course requires careful considerations of the instructional design. Structuring material, formulating assessments, and other instructional design aspects must be carefully considered. These considerations are detailed through the exploration of a case study implementing ALICE in a graduate level Systems Biology course.
△ Less
Submitted 4 May, 2017; v1 submitted 17 March, 2017;
originally announced March 2017.
