scholarly journals Learning in Volatile Environments with the Bayes Factor Surprise

2021 ◽  
pp. 1-72
Author(s):  
Vasiliki Liakoni ◽  
Alireza Modirshanechi ◽  
Wulfram Gerstner ◽  
Johanni Brea
Keyword(s):  
Message Passing ◽  
Bayes Factor ◽  
Sequence Length ◽  
Physiological Indicators ◽  
The Family ◽  
Current Belief ◽  
Observation Sequence ◽  
Exact Bayesian Inference ◽  
Theoretical Insight ◽  
Better Than

Surprise-based learning allows agents to rapidly adapt to nonstationary stochastic environments characterized by sudden changes. We show that exact Bayesian inference in a hierarchical model gives rise to a surprise-modulated trade-off between forgetting old observations and integrating them with the new ones. The modulation depends on a probability ratio, which we call the Bayes factor surprise, that tests the prior belief against the current belief. We demonstrate that in several existing approximate algorithms, the Bayes Factor Surprise modulates the rate of adaptation to new observations. We derive three novel surprise-based algorithms, one in the family of particle filters, one in the family of variational learning, and one in the family of message passing, that have constant scaling in observation sequence length and particularly simple update dynamics for any distribution in the exponential family. Empirical results show that these surprise-based algorithms estimate parameters better than alternative approximate approaches and reach levels of performance comparable to computationally more expensive algorithms. The Bayes Factor Surprise is related to but different from the Shannon Surprise. In two hypothetical experiments, we make testable predictions for physiological indicators that dissociate the Bayes factor surprise from the Shannon Surprise. The theoretical insight of casting various approaches as surprise-based learning, as well as the proposed online algorithms, may be applied to the analysis of animal and human behavior and to reinforcement learning in nonstationary environments.

Covey, The Negro Breaker.

2019 ◽  
Author(s):  
Frederick Douglass
Keyword(s):  
The Family ◽  
Better Than ◽  
The Way

Journey to my new master’s—meditations by the way—view of Covey’s residence—the family—the author’s awkwardness as a field hand—a cruel beating—why it was given—description of Covey—first adventure at ox driving—hair breadth escapes—ox and man alike property—Covey’s manner of proceeding to whip—hard labor better than the whip...

scholarly journals Simulation of the performance and scalability of message passing interface (MPI) communications of atmospheric models running on exascale supercomputers

2018 ◽  
Vol 11 (8) ◽  
pp. 3409-3426 ◽  
Author(s):  
Yongjun Zheng ◽  
Philippe Marguinaud
Keyword(s):  
Message Passing ◽  
Message Passing Interface ◽  
Grid Point ◽  
Routing Algorithm ◽  
Point Of View ◽  
Atmospheric Models ◽  
Innovative Methods ◽  
Spectral Models ◽  
Communication Time ◽  
Better Than

Abstract. In this study, we identify the key message passing interface (MPI) operations required in atmospheric modelling; then, we use a skeleton program and a simulation framework (based on SST/macro simulation package) to simulate these MPI operations (transposition, halo exchange, and allreduce), with the perspective of future exascale machines in mind. The experimental results show that the choice of the collective algorithm has a great impact on the performance of communications; in particular, we find that the generalized ring-k algorithm for the alltoallv operation and the generalized recursive-k algorithm for the allreduce operation perform the best. In addition, we observe that the impacts of interconnect topologies and routing algorithms on the performance and scalability of transpositions, halo exchange, and allreduce operations are significant. However, the routing algorithm has a negligible impact on the performance of allreduce operations because of its small message size. It is impossible to infinitely grow bandwidth and reduce latency due to hardware limitations. Thus, congestion may occur and limit the continuous improvement of the performance of communications. The experiments show that the performance of communications can be improved when congestion is mitigated by a proper configuration of the topology and routing algorithm, which uniformly distribute the congestion over the interconnect network to avoid the hotspots and bottlenecks caused by congestion. It is generally believed that the transpositions seriously limit the scalability of the spectral models. The experiments show that the communication time of the transposition is larger than those of the wide halo exchange for the semi-Lagrangian method and the allreduce in the generalized conjugate residual (GCR) iterative solver for the semi-implicit method below 2×105 MPI processes. The transposition whose communication time decreases quickly with increasing number of MPI processes demonstrates strong scalability in the case of very large grids and moderate latencies. The halo exchange whose communication time decreases more slowly than that of transposition with increasing number of MPI processes reveals its weak scalability. In contrast, the allreduce whose communication time increases with increasing number of MPI processes does not scale well. From this point of view, the scalability of spectral models could still be acceptable. Therefore it seems to be premature to conclude that the scalability of the grid-point models is better than that of spectral models at the exascale, unless innovative methods are exploited to mitigate the problem of the scalability presented in the grid-point models.

Kernels for Longitudinal Data with Variable Sequence Length and Sampling Intervals

2011 ◽  
Vol 23 (9) ◽  
pp. 2390-2420 ◽  
Author(s):  
Zhengdong Lu ◽  
Todd K. Leen ◽  
Jeffrey Kaye
Keyword(s):  
Longitudinal Data ◽  
Cognitive Decline ◽  
Likelihood Ratio ◽  
Neuropsychological Tests ◽  
Motor Behavior ◽  
Generative Models ◽  
Mixed Effects ◽  
Mixed Effects Models ◽  
Sequence Length ◽  
Better Than

We develop several kernel methods for classification of longitudinal data and apply them to detect cognitive decline in the elderly. We first develop mixed-effects models, a type of hierarchical empirical Bayes generative models, for the time series. After demonstrating their utility in likelihood ratio classifiers (and the improvement over standard regression models for such classifiers), we develop novel Fisher kernels based on mixture of mixed-effects models and use them in support vector machine classifiers. The hierarchical generative model allows us to handle variations in sequence length and sampling interval gracefully. We also give nonparametric kernels not based on generative models, but rather on the reproducing kernel Hilbert space. We apply the methods to detecting cognitive decline from longitudinal clinical data on motor and neuropsychological tests. The likelihood ratio classifiers based on the neuropsychological tests perform better than than classifiers based on the motor behavior. Discriminant classifiers performed better than likelihood ratio classifiers for the motor behavior tests.

scholarly journals Conhecimento das famílias de crianças usuárias de implante coclear sobre o desenvolvimento de linguagem e condições do dispositivo

2021 ◽  
Vol 47 (1) ◽  
Author(s):  
Natália Caroline Rovere ◽  
Arthur Menino Castilho ◽  
Maria Cecília Marconi Pinheiro Lima
Keyword(s):  
Language Development ◽  
Cochlear Implants ◽  
Public Hospital ◽  
Receptive Language ◽  
Deaf Children ◽  
Language Therapy ◽  
Telephone Contact ◽  
Number Of Patients ◽  
The Family ◽  
Better Than

RESUMOObjetivo: Conhecer como as famílias de crianças surdas usuárias de Implante Coclear percebem o desenvolvimento de linguagem da criança e relacionar esses dados com o tempo de uso diário e a quebra do dispositivo. Métodos: Trata-se de pesquisa de caráter exploratório com abordagem quantitativa. Realizou-se um levantamento do número de crianças implantadas em um serviço de saúde auditiva de um hospital escola, entre 2013 e 2015 e para aqueles que aceitaram participar da pesquisa, aplicou-se entrevista e a Escala de Aquisições Iniciais de Linguagem (ELM) com um dos membros da família, sendo esta áudio-gravada, via contato telefônico. Resultados: Das 66 crianças implantadas neste período, 52 participaram da pesquisa, sendo que destas 51 utilizavam o dispositivo. Verificou-se que a maioria (90,4%) frequentava terapia fonoaudiológica e as famílias relataram que estimulam o desenvolvimento da linguagem das crianças em casa. Aquelas classificadas com desenvolvimento de linguagem típico receberam o implante coclear precocemente, utilizavam o implante coclear o dia todo e nunca tiveram o equipamento quebrado. Quanto à Escala ELM, foi encontrado que existem mais casos de desenvolvimento típico para linguagem receptiva do que para a expressiva. Conclusão: Houve relação entre o uso do dispositivo e o desenvolvimento típico e a quebra do equipamento com desenvolvimento atípico de linguagem. Os familiares referiram que a compreensão de linguagem encontra-se mais adequada do que a produção da fala. Descritores: Implante coclear; Surdez; Perda auditiva; Linguagem; Desenvolvimento infantil ABSTRACT Objective: To know as the families of deaf children with cochlear implants (CI) perceive their language development and the to relate this data with the time of daily use and problems in the device. Methods: The research was observational, exploratory, with quantitative approach. A survey of the number of patients implanted in a Hearing Health Service of a public hospital was done, between 2013 and 2015 and it was applied an interview and the Early language Milestone Scale with one of the family members, audio-recorded, through telephone contact. Results: Out of the 66 implanted children, 52 families participated in the research and 51 used the device. It was verified that most of the children (90.04%) were in speech and language therapy and the families stimulated language development at home. The children with typical language development got the IC early in life, used the device all day and the implant never broke. In the ELM scale, it was found that there are more cases of typical development for receptive language than for expressive. Conclusion: There was a relationship between the use of the device and typical language development and the problems in the device with atypical language development. The families see that the children show receptive language development better than the expressive language.Key words: Cochlear Implantation; Deafness; Hearing Loss; Language; Child Development

scholarly journals Enhancement of chemical constituents in hydrosol and residual water of Aquilaria malaccensis tetraploid

2018 ◽  
Author(s):  
Siti Suhaila A Rahman ◽  
Keyword(s):  
Chemical Constituents ◽  
Water Soluble ◽  
Residual Water ◽  
Planting Material ◽  
Aquilaria Malaccensis ◽  
The Family ◽  
Material Choice ◽  
By Products ◽  
Better Than

Aquilaria malaccensis is an agarwood-producing species in the family Thymeleaeceae. Agarwood is a fragrant resin used in the manufacture of incense sticks, and in pharmaceutical, perfumery and cosmetic industries. In addition to the resin, hydrosol and residual water by-products from agarwood woodchip distillation are also utilized. Hydrosol contains water-soluble fragrant chemicals used as a tonic drink, in cooking and cosmetics while the residual water is used in spas and aromatic bath treatments. The present study was conducted to identify and compare compounds present in hydrosol and residual water by-products of diploid and polyploid A. malaccensis. Four different four-month-old A. malaccensis plants were compared: soil-grown diploid seedlings (DS), in vitro-grown seedlings (DV), tissue culture-derived plantlets (DC) and artificially induced tetraploid plantlets (TC). Hydrosol water from TC leaf and root samples were found to contain higher amounts of compounds compared with other samples. The TC leaf samples were qualitatively better as key compounds of agarwood such as α- and γ-eudesmol were detected. TC stem samples also contained higher amounts of key compounds compared with other samples, while the overall amount of compounds was highest in DS stem samples. The residual water of TC stem and root samples contained key compounds not detected in other samples, while DS residual water samples contained the highest total amount of compounds. Aquilaria malaccensis tetraploids performed better than their diploid counterparts in production of compounds, and thus may be a better planting material choice for commercial plantations.

scholarly journals Self-assembling Peptide Discovery: Overcoming Human Bias With Machine Learning

2021 ◽  
Author(s):  
Rohit Batra ◽  
Troy Loeffler ◽  
Henry Chan ◽  
Srilok Sriniva ◽  
Honggang Cui ◽  
...  
Keyword(s):  
Machine Learning ◽  
Self Assembly ◽  
Search Space ◽  
Sequence Length ◽  
Monte Carlo Tree Search ◽  
Self Assembling ◽  
Naturally Occurring ◽  
Peptide Materials ◽  
Dynamics Simulations ◽  
Better Than

Abstract Peptide materials have a wide array of functions from tissue engineering, surface coatings to catalysis and sensing. This class of biopolymer is composed of a sequence, comprised of 20 naturally occurring amino acids whose arrangement dictate the peptide functionality. While it is highly desirable to tailor the amino acid sequence, a small increase in their sequence length leads to dramatic increase in the possible candidates (e.g., from tripeptide = 20^3 or 8,000 peptides to a pentapeptide = 20^5 or 3.2 M). Traditionally, peptide design is guided by the use of structural propensity tables, hydrophobicity scales, or other desired properties and typically yields <10 peptides per study, barely scraping the surface of the search space. These approaches, driven by human expertise and intuition, are not easily scalable and are riddled with human bias. Here, we introduce a machine learning workflow that combines Monte Carlo tree search and random forest, with molecular dynamics simulations to develop a fully autonomous computational search engine (named, AI-expert) to discover peptide sequences with high potential for self-assembly (as a representative target functionality). We demonstrate the efficacy of the AI-expert to efficiently search large spaces of tripeptides and pentapeptides. Subsequent experiments on the proposed peptide sequences are performed to compare the predictability of the AI-expert with those of human experts. The AI performs on-par or better than human experts and suggests several non-intuitive sequences with high self-assembly propensity, outlining its potential to overcome human bias and accelerate peptide discovery.

scholarly journals The Centrality of Women in Moral Teaching in Yoruba Family System

2011 ◽  
Vol 9 (1) ◽  
Author(s):  
Yetunde A. Aluko ◽  
Oluwasegun D. Onobanjo ◽  
Nurudeen Alliyu
Keyword(s):  
Social Order ◽  
Family System ◽  
Well Being ◽  
Human Society ◽  
Law And Order ◽  
Feminine Gender ◽  
The Family ◽  
The Feminine ◽  
Better Than

Social order and peaceful co-existence are some of the primary goals in every human society. Central to maintenance of law and order in traditional Yoruba societies is the family. Culturally among the Yoruba people, women are socialized differently from men. This paper, rather than focus on the oppression of women in Yoruba cultural setting, examines the series of significant contributions of women to the maintenance of social order and ethical well-being of families. The feminine gender is not always synonymous with oppression and domination rather family well-being is mediated by the principle of complementarily between males and females. As a result, women have privileged relationship with their children better than men. Women are the pedagogues to lead their children, and this requires them to live a life worthy of emulation. The paper avers that the argument is not that men are totally left out of the moral upbringing of their children; most times they provide supportive roles. But the fact that women spend more time with the children calls for examination of their roles in the moral standing of children in Yoruba family system. Given the central role of women in the home and society, social order can be attained where women are committed to the moral upbringing of their children, rather than being marginalized in any aspect of the society.

scholarly journals Atașamentul securizant ca resursă a competenței parentale

2021 ◽  
Vol 39 (2) ◽  
pp. 21-26
Author(s):  
Lucia Gavriliţă
Keyword(s):  
Social Group ◽  
Parental Education ◽  
Social Behaviors ◽  
Emotional Security ◽  
Secure Attachment ◽  
Parental Competence ◽  
Attachment Patterns ◽  
The Family ◽  
Better Than

This study represents a system of experiments and data accompanied by scientifi c and methodological interpretations, of attachment patterns. Secure attachment is considered by many authors to be a source of parental competence. These quintessences express both, the conception of parental education and their attitude towards the typical child or the child with disabilities. Studies show that some parents adapt better than others to the child’s particularities. The family, specifi cally, off ers the child the climate of emotional security and is the fi rst social group in which the child practices social behaviors and discovers himself.

scholarly journals Use Of ACS 712ELC-5A Current Sensor on Overloaded Load Installation Safety System

2021 ◽  
Vol 4 (1) ◽  
pp. 47-55
Author(s):  
Anggara Trisna Nugraha Angga ◽  
Alwy Muhammad Ravi ◽  
Dadang Priyambodo ◽  
Adam Meredita Realdo ◽  
Fahmi Ivanuri ◽  
...  
Keyword(s):  
Safety System ◽  
Cutting Edge ◽  
Current Sensor ◽  
The Family ◽  
Protection Circuit ◽  
Edge Cost ◽  
Better Than

People want energy to satisfy their each day needs. However, maximum of those energy customers are now no longer aware that using an electric powered load is too large, which regularly reasons disturbances. In coping with overload disturbances withinside the family sector, MCB is used as a protection tool. The running precept of the MCB continues to be the usage of magnetic, so it's far less effective. When disconnecting the cutting-edge because of overload, the MCB will experience a touch overdue, which could reason sparks and a hearthplace withinside the family. Therefore, on this look at a cutting-edge sensor protection circuit is designed to update the MCB. This circuit can show the cutting-edge cost at the LCD and restriction the overcurrent to a restriction of three amperes. The overload safety machine constructed from the ACS712ELC-5A cutting-edge sensor can come across currents among 0A and 5A. So, to reduce off the cutting-edge, a 12V relay is used. When trying out a ready-made machine, this protection tool can study currents withinside the variety 0A to 3A, with median studying mistakes of 0.65%. If the burden is better than 3A, the machine also can reduce the cutting-edge.

The Unlikelihood of Family: A Photographic Essay on Transnational Experiences

2015 ◽  
Vol 85 (3) ◽  
pp. 342-356 ◽  
Author(s):  
Cristina Llerena Navarro
Keyword(s):  
United States ◽  
Immigration Policy ◽  
The United States ◽  
The Family ◽  
The Everyday ◽  
Photographic Essay ◽  
Photo Essay ◽  
Mixed Status Families ◽  
Mixed Status ◽  
Better Than

In this photo essay, Cristina Llerena Navarro captures moments in the everyday lives of mixed-status families. Through her narrative and images, Llerena shares the stories of these families, their journeys to the United States as well as the consequences of deportation on the family unity. She evokes the children's deep yearning to be reunited with their families on American soil, the parents' determination to provide their children with lives better than their own, and the realities of current immigration policy in preventing the fulfillment of these dreams.

Sign in / Sign up

Export Citation Format

Share Document