Open Access. Powered by Scholars. Published by Universities.®
Physical Sciences and Mathematics Commons™
Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 9 of 9
Full-Text Articles in Physical Sciences and Mathematics
Assessing The Costs Of Sampling Methods In Active Learning For Annotation, James Carroll, Robbie Haertel, Peter Mcclanahan, Eric K. Ringger, Kevin Seppi
Assessing The Costs Of Sampling Methods In Active Learning For Annotation, James Carroll, Robbie Haertel, Peter Mcclanahan, Eric K. Ringger, Kevin Seppi
Faculty Publications
Traditional Active Learning (AL) techniques assume that the annotation of each datum costs the same. This is not the case when annotating sequences; some sequences will take longer than others. We show that the AL technique which performs best depends on how cost is measured. Applying an hourly cost model based on the results of an annotation user study, we approximate the amount of time necessary to annotate a given sentence. This model allows us to evaluate the effectiveness of AL sampling methods in terms of time spent in annotation. We acheive a 77% reduction in hours from a random …
Learning Policies For Embodied Virtual Agents Through Demonstration, Jonathan Dinerstein, Parris K. Egbert, Dan A. Ventura
Learning Policies For Embodied Virtual Agents Through Demonstration, Jonathan Dinerstein, Parris K. Egbert, Dan A. Ventura
Faculty Publications
Although many powerful AI and machine learning techniques exist, it remains difficult to quickly create AI for embodied virtual agents that produces visually lifelike behavior. This is important for applications (e.g., games, simulators, interactive displays) where an agent must behave in a manner that appears human-like. We present a novel technique for learning reactive policies that mimic demonstrated human behavior. The user demonstrates the desired behavior by dictating the agent’s actions during an interactive animation. Later, when the agent is to behave autonomously, the recorded data is generalized to form a continuous state-to-action mapping. Combined with an appropriate animation algorithm …
A Data-Dependent Distance Measure For Transductive Instance-Based Learning, Jared Lundell, Dan A. Ventura
A Data-Dependent Distance Measure For Transductive Instance-Based Learning, Jared Lundell, Dan A. Ventura
Faculty Publications
We consider learning in a transductive setting using instance-based learning (k-NN) and present a method for constructing a data-dependent distance “metric” using both labeled training data as well as available unlabeled data (that is to be classified by the model). This new data-driven measure of distance is empirically studied in the context of various instance-based models and is shown to reduce error (compared to traditional models) under certain learning conditions. Generalizations and improvements are suggested.
Active Learning For Part-Of-Speech Tagging: Accelerating Corpus Annotation, George Busby, Marc Carmen, James Carroll, Robbie Haertel, Deryle W. Lonsdale, Peter Mcclanahan, Eric K. Ringger, Kevin Seppi
Active Learning For Part-Of-Speech Tagging: Accelerating Corpus Annotation, George Busby, Marc Carmen, James Carroll, Robbie Haertel, Deryle W. Lonsdale, Peter Mcclanahan, Eric K. Ringger, Kevin Seppi
Faculty Publications
In the construction of a part-of-speech annotated corpus, we are constrained by a fixed budget. A fully annotated corpus is required, but we can afford to label only a subset. We train a Maximum Entropy Markov Model tagger from a labeled subset and automatically tag the remainder. This paper addresses the question of where to focus our manual tagging efforts in order to deliver an annotation of highest quality. In this context, we find that active learning is always helpful. We focus on Query by Uncertainty (QBU) and Query by Committee (QBC) and report on experiments with several baselines and …
Particle Swarm Optimization In Dynamic Pricing, Christopher K. Monson, Patrick B. Mullen, Kevin Seppi, Sean C. Warnick
Particle Swarm Optimization In Dynamic Pricing, Christopher K. Monson, Patrick B. Mullen, Kevin Seppi, Sean C. Warnick
Faculty Publications
Dynamic pricing is a real-time machine learning problem with scarce prior data and a concrete learning cost. While the Kalman Filter can be employed to track hidden demand parameters and extensions to it can facilitate exploration for faster learning, the exploratory nature of Particle Swarm Optimization makes it a natural choice for the dynamic pricing problem. We compare both the Kalman Filter and existing particle swarm adaptations for dynamic and/or noisy environments with a novel approach that time-decays each particle's previous best value; this new strategy provides more graceful and effective transitions between exploitation and exploration, a necessity in the …
Task Similarity Measures For Transfer In Reinforcement Learning Task Libraries, James Carroll, Kevin Seppi
Task Similarity Measures For Transfer In Reinforcement Learning Task Libraries, James Carroll, Kevin Seppi
Faculty Publications
Recent research in task transfer and task clustering has necessitated the need for task similarity measures in reinforcement learning. Determining task similarity is necessary for selective transfer where only information from relevant tasks and portions of a task are transferred. Which task similarity measure to use is not immediately obvious. It can be shown that no single task similarity measure is uniformly superior. The optimal task similarity measure is dependent upon the task transfer method being employed. We define similarity in terms of tasks, and propose several possible task similarity measures, dT, dp, dQ, and dR which are based on …
A Bayesian Technique For Task Localization In Multiple Goal Markov Decision Processes, James Carroll, Kevin Seppi
A Bayesian Technique For Task Localization In Multiple Goal Markov Decision Processes, James Carroll, Kevin Seppi
Faculty Publications
In a reinforcement learning task library system for Multiple Goal Markov Decision Process (MGMDP), localization in the task space allows the agent to determine whether a given task is already in its library in order to exploit previously learned experience. Task localization in MGMDPs can be accomplished through a Bayesian approach, however a trivial approach fails when the rewards are not distributed normally. This can be overcome through our Bayesian Task Localization Technique (BTLT).
Using Permutations Instead Of Student’S T Distribution For P-Values In Paired-Difference Algorithm Comparisons, Tony R. Martinez, Joshua Menke
Using Permutations Instead Of Student’S T Distribution For P-Values In Paired-Difference Algorithm Comparisons, Tony R. Martinez, Joshua Menke
Faculty Publications
The paired-difference t-test is commonly used in the machine learning community to determine whether one learning algorithm is better than another on a given learning task. This paper suggests the use of the permutation test instead hecause it calculates the exact p-value instead of an estimate. The permutation test is also distribution free and the time complexity is trivial for the commonly used 10-fold cross-validation paired-difference test. Results of experiments on real-world problems suggest it is not uncommon to see the t-test estimate deviate up to 30-50% from the exact p-value.
Machine-Learned Contexts For Linguistic Operations In German Sentence Realization, Eric K. Ringger, Simon Corston-Oliver, Michael Gamon, Robert Moore
Machine-Learned Contexts For Linguistic Operations In German Sentence Realization, Eric K. Ringger, Simon Corston-Oliver, Michael Gamon, Robert Moore
Faculty Publications
We show that it is possible to learn the contexts for linguistic operations which map a semantic representation to a surface syntactic tree in sentence realization with high accuracy. We cast the problem of learning the contexts for the linguistic operations as classification tasks, and apply straightforward machine learning techniques, such as decision tree learning. The training data consist of linguistic features extracted from syntactic and semantic representations produced by a linguistic analysis system. The target features are extracted from links to surface syntax trees. Our evidence consists of four examples from the German sentence realization system code-named Amalgam: case …