Time-place learning in food-restricted Nile tilapia

Time-place learning based on food association was investigated in eight food-restricted Nile tilapias. Each fish was individually housed for 10 days in an experimental tank for adjustments to laboratory conditions, and fed daily in excess. Feeding was then interrupted for 17 days. Training was then started, based on a food-restricted regime in a tank divided into three interconnected compartments. Daily food was offered in one compartment (left or right side) of the tank in the morning and on the opposite side in the afternoon, for a continuous 30-day period. Frequency of choices on the right side was measured on days 10, 20 and 30 (during these test days, fish were not fed). Following this 30-day conditioning period, the Nile tilapias were able to switch sides at the correct period of the day to get food, suggesting that food restriction facilitates time-place learning discrimination. (C) 2007 Elsevier B.V. All rights reserved.

Testing time-place learning in the cichlid fish Nile tilapia

Time-place learning based on food association was investigated in the fish Nile tilapia. During a 30-day period, food was placed at one side of the aquarium (containing three compartments) in the morning and at the opposite side in the afternoon. Learning was inferred by the number of correct side choices of all fish in each day of test (15th, 30th). During the test day, fish were not fed. The Nile tilapia did not learn to switch sides at the correct day period in order to get food, suggesting thus that this species does not have time-place learning ability.

Time-place learning in individually reared angelfish, but not in pearl cichlid

Time-place learning based on food association was investigated in the cichlids angelfish (Pterophyllum scalare) and pearl cichlid (Geophagus brasiliensis) reared in isolation, therefore eliminating social influence on foraging. During a 30-day period, food was placed in one side of the aquarium (containing three compartments) in the morning and in the opposite side in the afternoon. Learning was inferred by the number of correct side choices of all fish in each day of test (15th and 30th). During the test day fish were not fed. The angelfish learned to switch sides at the correct day period in order to get food, suggesting this species has time-place learning ability when individually reared. on the other hand, the same was not observed for pearl cichlid. (c) 2006 Elsevier B.V. All rights reserved.

Influence of local environmental olfactory cues on place learning in rats

The aim of the present study was to assess the influence of local environmental olfactory cues on place learning in rats. We developed a new experimental design allowing the comparison of the use of local olfactory and visual cues in spatial and discrimination learning. We compared the effect of both types of cues on the discrimination of a single food source in an open-field arena. The goal was either in a fixed or in a variable location, and could be indicated by local olfactory and/or visual cues. The local cues enhanced the discrimination of the goal dish, whether it was in a fixed or in a variable location. However, we did not observe any overshadowing of the spatial information by the local olfactory or visual cue. Rats relied primarily on distant visuospatial information to locate the goal, neglecting local information when it was in conflict with the spatial information.

Semi-online neural-Q_leaming for real-time robot learning

Reinforcement learning (RL) is a very suitable technique for robot learning, as it can learn in unknown environments and in real-time computation. The main difficulties in adapting classic RL algorithms to robotic systems are the generalization problem and the correct observation of the Markovian state. This paper attempts to solve the generalization problem by proposing the semi-online neural-Q_learning algorithm (SONQL). The algorithm uses the classic Q_learning technique with two modifications. First, a neural network (NN) approximates the Q_function allowing the use of continuous states and actions. Second, a database of the most representative learning samples accelerates and stabilizes the convergence. The term semi-online is referred to the fact that the algorithm uses the current but also past learning samples. However, the algorithm is able to learn in real-time while the robot is interacting with the environment. The paper shows simulated results with the "mountain-car" benchmark and, also, real results with an underwater robot in a target following behavior

Indirect effects of peri-and postnatal choline treatment on place-learning abilities in rat.

This work was aimed at analyzing the effects of perinatal choline supplementation on the development of spatial abilities and upon adult performance. Choline supplementation (3.5 g/L in 0.02 M saccharin solution in tap water) was maintained for two weeks before birth and for up to four weeks postnatally. Additional supplementation was maintained from the fifth to the tenth week postnatally. Spatial-learning capacities were studied at the ages of 26, 65, or 80 days in a circular swimming pool (Morris place-navigation task) and at the age of 7 months in a homing arena. Treatment effects were found in both juvenile and adult rats, and thus persisted for several months after the cessation of the supplementation. The choline supplementation improved the performance in the water maze in a very selective manner. The most consistent effect was a reduction in the latency to reach a cued platform at a fixed position in space, whereas the improvement was limited when the platform was invisible and had to be located relative to distant cues only. However, after removal of the goal cue, the treated rats showed a better retention of the training position than did the control rats. A similar effect was observed in a dry-land task conducted in the homing arena. The choline supplementation thus induced a significant improvement of spatial memory. But since this effect was only evident following training with a salient cue, it might be regarded as an indirect effect promoted by an optimal combination of cue guidance with a place strategy.

Intrahippocampal cholinergic grafts in aged rats compensate impairments in a radial maze and in a place learning task

Age-related cognitive impairments were studied in rats kept in semi-enriched conditions during their whole life, and tested during ontogeny and adult life in various classical spatial tasks. In addition, the effect of intrahippocampal grafts of fetal septal-diagonal band tissue, rich in cholinergic neurons, was studied in some of these subjects. The rats received bilateral cell suspensions when aged 23-24 months. Starting 4 weeks after grafting, they were trained during 5 weeks in an 8-arm maze made of connected plexiglass tunnels. No age-related impairment was detected during the first eight trials, when the maze shape was that of a classical radial maze in which the rats had already been trained when young. The older rats were impaired when the task was made more difficult by rendering two arms parallel to each other. They developed an important neglect of one of the parallel tunnels resulting in a high amount of errors before completion of the task. In addition, the old rats developed a systematic response pattern of visits to adjacent arms in a sequence, which was not observed in the younger subjects. None of these behaviours were observed in the old rats with a septal transplant. Sixteen weeks after grafting, another experiment was conducted in a homing hole board task. Rats were allowed to escape from a large circular arena through one hole out of many, and to reach home via a flexible tube under the table. The escape hole was at a fixed position according to distant room cues, and olfactory cues were made irrelevant by rotating the table between the trials. An additional cue was placed on the escape position. No age-related difference in escape was observed during training. During a probe trial with no hole connected and no proximal cue present, the old untreated rats were less clearly focussed on the training sector than were either the younger or the grafted old subjects. Taken together, these experiments indicate that enriched housing conditions and spatial training during adult life do not protect against all age-related deterioration in spatial ability. However, it might be that the considerable improvement observed in the grafted subjects results from an interaction between the graft treatment and the housing conditions.

Quality of Learners' Time and Learning Performance Beyond Quantitative Time-on-Task

Peer-reviewed

Semi-online neural-Q_leaming for real-time robot learning

Reinforcement learning (RL) is a very suitable technique for robot learning, as it can learn in unknown environments and in real-time computation. The main difficulties in adapting classic RL algorithms to robotic systems are the generalization problem and the correct observation of the Markovian state. This paper attempts to solve the generalization problem by proposing the semi-online neural-Q_learning algorithm (SONQL). The algorithm uses the classic Q_learning technique with two modifications. First, a neural network (NN) approximates the Q_function allowing the use of continuous states and actions. Second, a database of the most representative learning samples accelerates and stabilizes the convergence. The term semi-online is referred to the fact that the algorithm uses the current but also past learning samples. However, the algorithm is able to learn in real-time while the robot is interacting with the environment. The paper shows simulated results with the "mountain-car" benchmark and, also, real results with an underwater robot in a target following behavior

An investigation of two real time machine learning techniques that could enhance the adaptability of game AI agents

Developers strive to create innovative Artificial Intelligence (AI) behaviour in their games as a key selling point. Machine Learning is an area of AI that looks at how applications and agents can be programmed to learn their own behaviour without the need to manually design and implement each aspect of it. Machine learning methods have been utilised infrequently within games and are usually trained to learn offline before the game is released to the players. In order to investigate new ways AI could be applied innovatively to games it is wise to explore how machine learning methods could be utilised in real-time as the game is played, so as to allow AI agents to learn directly from the player or their environment. Two machine learning methods were implemented into a simple 2D Fighter test game to allow the agents to fully showcase their learned behaviour as the game is played. The methods chosen were: Q-Learning and an NGram based system. It was found that N-Grams and QLearning could significantly benefit game developers as they facilitate fast, realistic learning at run-time.

Aprendizagem espaço-temporal e efeitos das condições luminosas no aprendizado do peixe paulistinha

Para processar a informação ambiental e perceber o tempo, os indivíduos utilizam-se de pistas ambientais, como luz e temperatura, que servem como guias para o relógio interno. O mecanismo temporizador endógeno é chamado relógio circadiano, o qual comanda uma grande variedade de ritmos diários bioquímicos, fisiológicos e comportamentais presentes nos organismos. Com isso, os animais podem antecipar eventos espaço-temporalmente distribuídos e usar essa informação para organizar as atividades diárias, o que é uma vantagem adaptativa para os indivíduos, já que muitos fatores ambientais apresentam variação circadiana. Aprendizagem espaço-temporal (do inglês: "time-place learning’’-TPL) é a habilidade de associar lugares com importantes eventos biológicos em diferentes horas do dia. Em nosso estudo utilizamos como modelo o peixe paulistinha (Danio rerio), conhecido por ser altamente social, para testar aprendizagem espaço-temporal baseada em reforço social. Além disso, objetivamos averiguar os efeitos das condições de claro constante e escuro constante na aprendizagem espaço-temporal, e se nessas condições, a atividade do peixe paulistinha é alterada. Para isso, testamos três diferentes condições (n=10): grupo claro-escuro (CE), grupo claro constante (CC) e grupo escuro constante (EE) durante 30 dias da seguinte maneira: diariamente, um grupo de 5 peixes paulistinha foi introduzido em um recipiente localizado no compartimento da manhã (um dos lados do aquário), às 8:00h e retirado às 9:00h, e em outro recipiente do compartimento da tarde (lado oposto do aquário), às 17:00h e removido às 18:00h, servindo como estímulo para que o peixe experimental ocupasse o compartimento onde o grupo fosse colocado. O comportamento foi filmado nos dois horários, 15 minutos antes e durante os 60 minutos de exposição ao estímulo, no 15º e no 30ª dia, porém neste último, os peixes foram filmados sem a presença do estímulo a fim de averiguarmos a aprendizagem espaço-temporal. Por fim, para saber a influência das três condições luminosas na atividade dos peixes, filmamos os últimos 6 dias de teste, para registrar o padrão de atividade. Nossos resultados mostraram que em ciclo claro-escuro (CE) o peixe paulistinha apresenta TPL, bem como é capaz de antecipar a hora e local do estímulo (grupo de coespecíficos), enfatizando a importância do estímulo social para a aprendizagem. Em condições de claro constante e escuro constante, o peixe paulistinha não apresentou aprendizagem espaço-temporal. Ademais, após 30 dias em condições luminosas constantes (claro constante e escuro constante), o peixe paulistinha mantém ritmo circadiano, porém em claro constante sua atividade é aumentada e seu ritmo atividade-repouso é alterado, através de um padrão de atividade distribuída homogeneamente ao longo das 24h, ao invés de concentrada na subjetiva fase clara, como nos grupos de ciclo claro-escuro e escuro constante, os quais conservam o padrão de atividade diurno da espécie.

Development of place navigation in rats from weaning to puberty.

Young hooded rats were trained to escape onto a hidden platform after swimming in a pool of opaque water. Subjects 21, 28, 35, 42, and 64 days of age on the first training day were given 28 trials on 5 consecutive days. Half of the rats were required to localize the platform in relation to external room cues only ("place only" condition) and the other half were helped by the presence of a visible cue on the platform ("cue + place" condition). A deficiency in place navigation was observed in the 21- and 28-day groups; they showed slow escape and took circuitous routes more often than older rats. This deficiency was related to a poor spatial bias toward the training position when the subjects were allowed to swim for 30 s in the absence of the platform, at the end of the 28-trial training period (probe trial). The 35-day group showed adult-like learning ability in both training conditions, but failed to show searching behavior during the probe trial after having been trained in the presence of the proximal cue. Only rats older than 40 days showed typical adult behavior such as swimming directly toward the platform from any starting position and localized searching around the absent platform's position during the probe trial, no matter what the training conditions were. These results suggest that central nervous system structures responsible for place learning in the rat are functional from around 32 days of age, but fail to trigger searching behavior following cued training before the sixth week.