Bayesian Network Classifiers
... share the same structure G, and if Θ satisfies Equation 5, then LL(B|D) ≥ LL(B 0 |D). Thus, given a network structure, there is a closed form solution for the parameters that maximize the log likelihood score, namely, Equation 5. Moreover, since the first term of Equation 2 does not depend on the ch ...
... share the same structure G, and if Θ satisfies Equation 5, then LL(B|D) ≥ LL(B 0 |D). Thus, given a network structure, there is a closed form solution for the parameters that maximize the log likelihood score, namely, Equation 5. Moreover, since the first term of Equation 2 does not depend on the ch ...
Learning place cells, grid cells and invariances: A unifying model
... synaptic changes during spatial exploration. In principle, the time scale of plasticitybased models can be augmented arbitrarily by increasing the synaptic learning rates. For stable patterns to emerge, however, significant weight changes must occur only after the animal has visited most of the envi ...
... synaptic changes during spatial exploration. In principle, the time scale of plasticitybased models can be augmented arbitrarily by increasing the synaptic learning rates. For stable patterns to emerge, however, significant weight changes must occur only after the animal has visited most of the envi ...
Condition interference in rats performing a choice task with switched
... upon responses to any situation is essential. To date, however, decision processes and the underlying neural substrates have been investigated under specific conditions; thus, little is known about how various conditions influence one another in these processes. In this study, we designed a binary c ...
... upon responses to any situation is essential. To date, however, decision processes and the underlying neural substrates have been investigated under specific conditions; thus, little is known about how various conditions influence one another in these processes. In this study, we designed a binary c ...
Introduction to Artificial Intelligence (Undergraduate Topics in
... The fields of image processing, fuzzy logic, and natural language processing are not covered in detail. The field of image processing, which is important for all of computer science, is a stand-alone discipline with very good textbooks, such as [GW08]. Natural language processing has a similar statu ...
... The fields of image processing, fuzzy logic, and natural language processing are not covered in detail. The field of image processing, which is important for all of computer science, is a stand-alone discipline with very good textbooks, such as [GW08]. Natural language processing has a similar statu ...
Parallel contributions of distinct human memory systems during
... Regions within the medial temporal lobe and basal ganglia are thought to subserve distinct memory systems underlying declarative and nondeclarative processes, respectively. One question of interest is how these multiple memory systems interact during learning to contribute to goal directed behavior. ...
... Regions within the medial temporal lobe and basal ganglia are thought to subserve distinct memory systems underlying declarative and nondeclarative processes, respectively. One question of interest is how these multiple memory systems interact during learning to contribute to goal directed behavior. ...
CNS learns Stable, Accurate and Efficient Movements using a
... (Franklin et al., 2003a; Milner and Franklin, 2005). We identified common features in the evolution of muscle activation whether the force field elicited a stable or unstable interaction with the arm, which are captured in the three principles described below. If a muscle was stretched relative to i ...
... (Franklin et al., 2003a; Milner and Franklin, 2005). We identified common features in the evolution of muscle activation whether the force field elicited a stable or unstable interaction with the arm, which are captured in the three principles described below. If a muscle was stretched relative to i ...
Inverse Reinforcement Learning in Relational Domains
... The blocks world offers two abstract actions move(X, Y ) and wait(). The wait() abstract action is available in every state and does not modify the state. The move(X, Y ) abstract action is defined with three rules that describe when the actions can be applied and what becomes the next state: for ex ...
... The blocks world offers two abstract actions move(X, Y ) and wait(). The wait() abstract action is available in every state and does not modify the state. The move(X, Y ) abstract action is defined with three rules that describe when the actions can be applied and what becomes the next state: for ex ...
Implications of Polychronous Neuronal Groups for the Nature of Mental Representations
... 1, respectively, providing it with no coincident spikes to drive an action potential. Alternatively, if neurons a, b, and c fire in the reverse order, neuron y will may spike, while neuron x will remain silent. Thus, the effects of spikes from neurons a, b, and c on the firing of neurons x and y is ...
... 1, respectively, providing it with no coincident spikes to drive an action potential. Alternatively, if neurons a, b, and c fire in the reverse order, neuron y will may spike, while neuron x will remain silent. Thus, the effects of spikes from neurons a, b, and c on the firing of neurons x and y is ...
Stimulus Configuration, Classical Conditioning, and
... cells is positively correlated with the topography of the CR. Because one important hippocampal output is mediated through CAS axons that reach the lateral septum, it is not surprising that Berger and Thompson (1978b) and Salvatierra and Berry (1989) reported that neural activity in the lateral sept ...
... cells is positively correlated with the topography of the CR. Because one important hippocampal output is mediated through CAS axons that reach the lateral septum, it is not surprising that Berger and Thompson (1978b) and Salvatierra and Berry (1989) reported that neural activity in the lateral sept ...
A model for experience-dependent changes in the responses of inferotemporal neurons
... Abstract. Neurons in inferior temporal (IT) cortex exhibit selectivity for complex visual stimuli and can maintain activity during the delay following the presentation of a stimulus in delayed match to sample tasks. Experimental work in awake monkeys has shown that the responses of IT neurons declin ...
... Abstract. Neurons in inferior temporal (IT) cortex exhibit selectivity for complex visual stimuli and can maintain activity during the delay following the presentation of a stimulus in delayed match to sample tasks. Experimental work in awake monkeys has shown that the responses of IT neurons declin ...
Learning logical definitions from relations
... task. A fixed number of attribute values conveys a fixed number of bits of information and it is easy to construct a network that would require more bits to represent its nodes and links. Further, suppose the description task were simplified by restricting networks to a maximum of ten nodes, say, wi ...
... task. A fixed number of attribute values conveys a fixed number of bits of information and it is easy to construct a network that would require more bits to represent its nodes and links. Further, suppose the description task were simplified by restricting networks to a maximum of ten nodes, say, wi ...
Following non-stationary distributions by controlling the
... of examples, chosen from PX , and minimizes the distortion measured on this set [15,11]. Some other methods works on-line, since examples are provided continuously. In this latter case, at each stage, an input ξ is first chosen, according to PX . Second, a so called winner-take-all procedure (WTA) a ...
... of examples, chosen from PX , and minimizes the distortion measured on this set [15,11]. Some other methods works on-line, since examples are provided continuously. In this latter case, at each stage, an input ξ is first chosen, according to PX . Second, a so called winner-take-all procedure (WTA) a ...
ConceptNet - Media Lab Login
... hierarchical relations. Its most recent version 2.0 contains roughly 200,000 word “senses” (a sense is a “distinct” meaning that a word can assume). One of the reasons for its success and wide adoption is its ease of use. As a simple semantic network with words at the nodes, it can be readily applie ...
... hierarchical relations. Its most recent version 2.0 contains roughly 200,000 word “senses” (a sense is a “distinct” meaning that a word can assume). One of the reasons for its success and wide adoption is its ease of use. As a simple semantic network with words at the nodes, it can be readily applie ...
review - NYU Psychology
... act as an US, but only when the observer believes that it was caused by a shock, not when the model’s arm moves without a shock or when a shock is delivered without arm movements37. These results support the conclusion that perceptual properties of the learning model interact with the observer’s kno ...
... act as an US, but only when the observer believes that it was caused by a shock, not when the model’s arm moves without a shock or when a shock is delivered without arm movements37. These results support the conclusion that perceptual properties of the learning model interact with the observer’s kno ...
Mind Design II : Philosophy, Psychology, Artificial Intelligence
... than is traditional empirical psychology. An "experiment" in mind design is more often an effort to build something and make it work, than to observe or analyze what already exists. Thus, the field of artificial intelligence (AI), the attempt to construct intelligent artifacts, systems with minds of ...
... than is traditional empirical psychology. An "experiment" in mind design is more often an effort to build something and make it work, than to observe or analyze what already exists. Thus, the field of artificial intelligence (AI), the attempt to construct intelligent artifacts, systems with minds of ...
Mind Design II : Philosophy, Psychology, Artificial Intelligence
... than is traditional empirical psychology. An "experiment" in mind design is more often an effort to build something and make it work, than to observe or analyze what already exists. Thus, the field of artificial intelligence (AI), the attempt to construct intelligent artifacts, systems with minds of ...
... than is traditional empirical psychology. An "experiment" in mind design is more often an effort to build something and make it work, than to observe or analyze what already exists. Thus, the field of artificial intelligence (AI), the attempt to construct intelligent artifacts, systems with minds of ...
VALUE-DEPENDENT SELECTION IN THE BRAIN: SIMULATION IN
... the improvement that ensues when the connections to the value system are themselves plastic and thus become able to mediate acquired value. Using a second-order conditioning paradigm, we demonstrate that auditory discrimination can occur in the model in the absence of direct positive reinforcement a ...
... the improvement that ensues when the connections to the value system are themselves plastic and thus become able to mediate acquired value. Using a second-order conditioning paradigm, we demonstrate that auditory discrimination can occur in the model in the absence of direct positive reinforcement a ...
Neural Crest - bthsresearch
... • If N-cadherin is overexpressed in the surrounding surface ectoderm, neural tube closure is impeded – This is achieved by injecting N-cadherin mRNA into the embryo at the 2-cell stage ...
... • If N-cadherin is overexpressed in the surrounding surface ectoderm, neural tube closure is impeded – This is achieved by injecting N-cadherin mRNA into the embryo at the 2-cell stage ...
sequential decision models for expert system optimization
... Value maximization research, on the other hand, does not enforce a consistency requirement. The knowledge source is used to extract the probability of a class or a future outcome, conditioned upon certain input observations. In value maximization problems, decision rules are typically constructed us ...
... Value maximization research, on the other hand, does not enforce a consistency requirement. The knowledge source is used to extract the probability of a class or a future outcome, conditioned upon certain input observations. In value maximization problems, decision rules are typically constructed us ...
Lecture notes Neural Computation
... The brain is a complex computing machine which is evolving to give the “fittest” output to a given input. Neural computation has as goal to describe the function of the nervous system in mathematical terms. By analysing or simulating the resulting equations, one can better understand its function, r ...
... The brain is a complex computing machine which is evolving to give the “fittest” output to a given input. Neural computation has as goal to describe the function of the nervous system in mathematical terms. By analysing or simulating the resulting equations, one can better understand its function, r ...
Catastrophic interference
Catastrophic Interference, also known as catastrophic forgetting, is the tendency of a artificial neural network to completely and abruptly forget previously learned information upon learning new information. Neural networks are an important part of the network approach and connectionist approach to cognitive science. These networks use computer simulations to try and model human behaviours, such as memory and learning. Catastrophic interference is an important issue to consider when creating connectionist models of memory. It was originally brought to the attention of the scientific community by research from McCloskey and Cohen (1989), and Ractcliff (1990). It is a radical manifestation of the ‘sensitivity-stability’ dilemma or the ‘stability-plasticity’ dilemma. Specifically, these problems refer to the issue of being able to make an artificial neural network that is sensitive to, but not disrupted by, new information. Lookup tables and connectionist networks lie on the opposite sides of the stability plasticity spectrum. The former remains completely stable in the presence of new information but lacks the ability to generalize, i.e. infer general principles, from new inputs. On the other hand, connectionst networks like the standard backpropagation network are very sensitive to new information and can generalize on new inputs. Backpropagation models can be considered good models of human memory insofar as they mirror the human ability to generalize but these networks often exhibit less stability than human memory. Notably, these backpropagation networks are susceptible to catastrophic interference. This is considered an issue when attempting to model human memory because, unlike these networks, humans typically do not show catastrophic forgetting. Thus, the issue of catastrophic interference must be eradicated from these backpropagation models in order to enhance the plausibility as models of human memory.