
Code-specific policy gradient rules for spiking neurons
... partially observable Markov decision processes, the input spike trains are observations that provide information about the state of the animal and the output spike trains are controls that influence the action choice. Depending on both of these spike trains, the system receives a reward. The goal is ...
... partially observable Markov decision processes, the input spike trains are observations that provide information about the state of the animal and the output spike trains are controls that influence the action choice. Depending on both of these spike trains, the system receives a reward. The goal is ...
Episodic memory as a prerequisite for online updates
... liberates us from the need to retain the whole data set: once the posterior has been updated the data can be discarded. As long as both parameters and models are updated, this procedure provides a consistent method to update and compare alternative hypotheses on how the model was generated without n ...
... liberates us from the need to retain the whole data set: once the posterior has been updated the data can be discarded. As long as both parameters and models are updated, this procedure provides a consistent method to update and compare alternative hypotheses on how the model was generated without n ...
Preference Learning with Gaussian Processes
... classification problem. Based on our recent work on ordinal regression (Chu & Ghahramani, 2004), we further develop the Gaussian process algorithm for preference learning tasks. Although the basic techniques we used in these two works are similar, the formulation proposed in this paper is new, more g ...
... classification problem. Based on our recent work on ordinal regression (Chu & Ghahramani, 2004), we further develop the Gaussian process algorithm for preference learning tasks. Although the basic techniques we used in these two works are similar, the formulation proposed in this paper is new, more g ...
Sources of Evidence-of-Learning: Learning and assessment in the
... emails. In the era of interpersonal computing, the social relations of information and communication can be systematically and consistently ordered. This opens out the social phenomenon that is popularly characterized as ‘Web 2.0’ (O’Reilly, 2005), one aspect of which is massively integrated social ...
... emails. In the era of interpersonal computing, the social relations of information and communication can be systematically and consistently ordered. This opens out the social phenomenon that is popularly characterized as ‘Web 2.0’ (O’Reilly, 2005), one aspect of which is massively integrated social ...
Topic 4
... learning to identify the right answer through an iterative process of self-adaptation or training If there are many factors, with complex interactions among them, the usual "linear" statistical techniques may be inappropriate If sufficient data is available, an ANN can find the relevant function ...
... learning to identify the right answer through an iterative process of self-adaptation or training If there are many factors, with complex interactions among them, the usual "linear" statistical techniques may be inappropriate If sufficient data is available, an ANN can find the relevant function ...
CIS 830: Advanced Topics in Artificial Intelligence KSU When
... – Solution: use networks of perceptrons (LTUs) ...
... – Solution: use networks of perceptrons (LTUs) ...
Multi-objective optimization of support vector machines
... Summary. Designing supervised learning systems is in general a multi-objective optimization problem. It requires finding appropriate trade-offs between several objectives, for example between model complexity and accuracy or sensitivity and specificity. We consider the adaptation of kernel and regul ...
... Summary. Designing supervised learning systems is in general a multi-objective optimization problem. It requires finding appropriate trade-offs between several objectives, for example between model complexity and accuracy or sensitivity and specificity. We consider the adaptation of kernel and regul ...
Kenji Doya 2001
... licate reward-based learning behaviors [11], [12]. Thus, by combining the experimental data Reward r(t) Action u(t) from the basal ganglia and the theory of reinforceFigure 5. A schematic diagram of the circuit of the basal ganglia and their loop ment learning, the role of the basal ganglia has beco ...
... licate reward-based learning behaviors [11], [12]. Thus, by combining the experimental data Reward r(t) Action u(t) from the basal ganglia and the theory of reinforceFigure 5. A schematic diagram of the circuit of the basal ganglia and their loop ment learning, the role of the basal ganglia has beco ...
Where is Education Heading and How About AI?
... problem-solving path thus generated to an ideal path - based on expert performance. In general, most ITS systems cope well with situations in which the student’s answer differs from that of the expert model. If a remediation move is needed, this takes the form of an explanation of what step the expe ...
... problem-solving path thus generated to an ideal path - based on expert performance. In general, most ITS systems cope well with situations in which the student’s answer differs from that of the expert model. If a remediation move is needed, this takes the form of an explanation of what step the expe ...
- White Rose Research Online
... the bagging prediction models on imbalanced data-sets. Most research on existing bagging-based sampling schemes for imbalanced data, e.g. (Li 2007; Hido, Kashima, and Takahashi 2009), focused on using sampling methods to provide a set of equally balanced or average-balanced training sub-sets for tra ...
... the bagging prediction models on imbalanced data-sets. Most research on existing bagging-based sampling schemes for imbalanced data, e.g. (Li 2007; Hido, Kashima, and Takahashi 2009), focused on using sampling methods to provide a set of equally balanced or average-balanced training sub-sets for tra ...
MS PowerPoint 97/2000 format
... – It uses scoring measure instead of mutual information to measure the dependency of parent and children, then uses the maximum score to build BBN – This algorithm can allow children to have multiple parents and handle random variables with multiple values. – The limited candidate sets provide a sma ...
... – It uses scoring measure instead of mutual information to measure the dependency of parent and children, then uses the maximum score to build BBN – This algorithm can allow children to have multiple parents and handle random variables with multiple values. – The limited candidate sets provide a sma ...
Introduction to AI (COMP-424) - McGill School Of Computer Science
... “We propose that a 2 month, 10 man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire. The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be s ...
... “We propose that a 2 month, 10 man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire. The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be s ...
AI AND MACHINE LEARNING TECHNIQUES FOR MANAGING
... sense that they enable the system to do the same task or tasks drawn from the same population more effectively the next time”. With respect to advanced engineering automation, as stated in (Lu, 1990), “we need new computer technologies that cannot only generate, record, and retrieve information, but ...
... sense that they enable the system to do the same task or tasks drawn from the same population more effectively the next time”. With respect to advanced engineering automation, as stated in (Lu, 1990), “we need new computer technologies that cannot only generate, record, and retrieve information, but ...
pdf - laral
... In this section we describe the experimental set-up used to perform our experiments and the learning algorithm used to develop experienced agents, i.e. agents able to correctly perform a foraging task. These agents will be used, in the experiments described in the following sections, to study whethe ...
... In this section we describe the experimental set-up used to perform our experiments and the learning algorithm used to develop experienced agents, i.e. agents able to correctly perform a foraging task. These agents will be used, in the experiments described in the following sections, to study whethe ...
Semi-Supervised Learning Using Gaussian Fields and Harmonic
... The semi-supervised learning problem has attracted an increasing amount of interest recently, and several novel approaches have been proposed; we refer to (Seeger, 2001) for an overview. Among these methods is a promising family of techniques that exploit the “manifold structure” of the data; such m ...
... The semi-supervised learning problem has attracted an increasing amount of interest recently, and several novel approaches have been proposed; we refer to (Seeger, 2001) for an overview. Among these methods is a promising family of techniques that exploit the “manifold structure” of the data; such m ...
Joint Regression and Linear Combination of Time
... this paper, is solved using a polynomial solver that uses solely basic linear algebra tools. Solving multivariate polynomial systems is normally done in the field of computational algebraic geometry [3]. Another class of polynomial solvers are homotopy continuation methods. These are a symbolic-nume ...
... this paper, is solved using a polynomial solver that uses solely basic linear algebra tools. Solving multivariate polynomial systems is normally done in the field of computational algebraic geometry [3]. Another class of polynomial solvers are homotopy continuation methods. These are a symbolic-nume ...
Machine learning

Machine learning is a subfield of computer science that evolved from the study of pattern recognition and computational learning theory in artificial intelligence. Machine learning explores the study and construction of algorithms that can learn from and make predictions on data. Such algorithms operate by building a model from example inputs in order to make data-driven predictions or decisions, rather than following strictly static program instructions.Machine learning is closely related to and often overlaps with computational statistics; a discipline that also specializes in prediction-making. It has strong ties to mathematical optimization, which delivers methods, theory and application domains to the field. Machine learning is employed in a range of computing tasks where designing and programming explicit algorithms is infeasible. Example applications include spam filtering, optical character recognition (OCR), search engines and computer vision. Machine learning is sometimes conflated with data mining, although that focuses more on exploratory data analysis. Machine learning and pattern recognition ""can be viewed as two facets ofthe same field.""When employed in industrial contexts, machine learning methods may be referred to as predictive analytics or predictive modelling.