Using Semi-Supervised Learning to Build Bayesian Network for Personal Preference Modeling in Home Environment

Analysis of a Method Improving Reinforcement Learning Agents’ Policies

Journal of Advanced Computational Intelligence and Intelligent Informatics ◽

10.20965/jaciii.2003.p0276 ◽

2003 ◽

Vol 7 (3) ◽

pp. 276-282 ◽

Cited By ~ 1

Author(s):

Daisuke Kitakoshi ◽

◽

Hiroyuki Shioya ◽

Masahito Kurihara ◽

Keyword(s):

Machine Learning ◽

Reinforcement Learning ◽

Stochastic Model ◽

Bayesian Network ◽

Supervised Learning ◽

Minimum Description Length ◽

Input Output ◽

Pursuit Problem ◽

Learning Agents ◽

Sample Data

Reinforcement learning (RL) is a kind of machine learning. It aims to optimize agents’ policies by adapting the agents to an environment according to rewards. In this paper, we propose a method for improving policies by using stochastic knowledge, in which reinforcement learning agents obtain. We use a Bayesian Network (BN), which is a stochastic model, as knowledge of an agent. Its structure is decided by minimum description length criterion using series of an agent’s input-output and rewards as sample data. A BN constructed in our study represents stochastic dependences between input-output and rewards. In our proposed method, policies are improved by supervised learning using the structure of BN (i.e. stochastic knowledge). The proposed improvement mechanism makes RL agents acquire more effective policies. We carry out simulations in the pursuit problem in order to show the effectiveness of our proposed method.

Download Full-text

Multi-variable flood damage modelling with limited data using supervised learning approaches

Natural Hazards and Earth System Science ◽

10.5194/nhess-17-1683-2017 ◽

2017 ◽

Vol 17 (9) ◽

pp. 1683-1696 ◽

Cited By ~ 28

Author(s):

Dennis Wagenaar ◽

Jurjen de Jong ◽

Laurens M. Bouwer

Keyword(s):

Bayesian Network ◽

Supervised Learning ◽

Water Depth ◽

Damage Assessment ◽

Regression Trees ◽

Flood Damage ◽

Data Set ◽

Widespread Application ◽

Learning Techniques ◽

Flood Damage Assessment

Abstract. Flood damage assessment is usually done with damage curves only dependent on the water depth. Several recent studies have shown that supervised learning techniques applied to a multi-variable data set can produce significantly better flood damage estimates. However, creating and applying a multi-variable flood damage model requires an extensive data set, which is rarely available, and this is currently holding back the widespread application of these techniques. In this paper we enrich a data set of residential building and contents damage from the Meuse flood of 1993 in the Netherlands, to make it suitable for multi-variable flood damage assessment. Results from 2-D flood simulations are used to add information on flow velocity, flood duration and the return period to the data set, and cadastre data are used to add information on building characteristics. Next, several statistical approaches are used to create multi-variable flood damage models, including regression trees, bagging regression trees, random forest, and a Bayesian network. Validation on data points from a test set shows that the enriched data set in combination with the supervised learning techniques delivers a 20 % reduction in the mean absolute error, compared to a simple model only based on the water depth, despite several limitations of the enriched data set. We find that with our data set, the tree-based methods perform better than the Bayesian network.

Download Full-text

Comparisons of Auditory Brainstem Responses Between a Laboratory and Simulated Home Environment

Journal of Speech Language and Hearing Research ◽

10.1044/2020_jslhr-20-00383 ◽

2020 ◽

Vol 63 (11) ◽

pp. 3877-3892

Author(s):

Ashley Parker ◽

Candace Slack ◽

Erika Skoe

Keyword(s):

Home Environment ◽

Signal To Noise Ratio ◽

Auditory Brainstem ◽

Auditory Brainstem Responses ◽

Single Family ◽

Adult Participants ◽

Simulated Home Environment ◽

Brainstem Response ◽

Research Populations ◽

Response Consistency

Purpose Miniaturization of digital technologies has created new opportunities for remote health care and neuroscientific fieldwork. The current study assesses comparisons between in-home auditory brainstem response (ABR) recordings and recordings obtained in a traditional lab setting. Method Click-evoked and speech-evoked ABRs were recorded in 12 normal-hearing, young adult participants over three test sessions in (a) a shielded sound booth within a research lab, (b) a simulated home environment, and (c) the research lab once more. The same single-family house was used for all home testing. Results Analyses of ABR latencies, a common clinical metric, showed high repeatability between the home and lab environments across both the click-evoked and speech-evoked ABRs. Like ABR latencies, response consistency and signal-to-noise ratio (SNR) were robust both in the lab and in the home and did not show significant differences between locations, although variability between the home and lab was higher than latencies, with two participants influencing this lower repeatability between locations. Response consistency and SNR also patterned together, with a trend for higher SNRs to pair with more consistent responses in both the home and lab environments. Conclusions Our findings demonstrate the feasibility of obtaining high-quality ABR recordings within a simulated home environment that closely approximate those recorded in a more traditional recording environment. This line of work may open doors to greater accessibility to underserved clinical and research populations.

Download Full-text