arXiv subjects

Guiliang Liu

Publications and source records attributed to Guiliang Liu.

40 records · Page 3Linked to original sources

arXiv2022

Asymmetric phenomenon of flow and heat transfer in charging process of thermal energy storage based on an entire domain model

Phase change energy storage is getting increasing attention as a representative technology to achieve carbon neutrality. The phase change process exists typical phenomenon of asymmetry that affects the energy storage performance. However, the mechanism of asymmetry is currently lack of elaboration because the half domain model is always used to simplify the numerical simulation and avoid the appearance of asymmetry. In this study, the entire domain model and boundary conditions were adopted and numerically simulated for the melting process of paraffin wax, i.e., the charging process of energy storage. The nonlinear dynamics method was applied to explain the asymmetric flow and heat transfer phenomenon. The three-stage characteristic based on charging speed was proposed for charging process and was explained by thermal conduction or natural convection. The asymmetric phenomenon was elaborated on the cause of formation, change mechanism, and effect evaluation. Results show that natural convection accounts for the multiple solutions (including the asymmetric ones) of charging process. It is also pointed out that asymmetric solutions can exist under symmetric geometry structure. To accurately solve the charging process, it is necessary to use the entire domain model. Both the charging speed and energy storage capacity have a positive correlation with asymmetry. Thus, a potential idea for energy storage application was proposed to increase the charging speed; that is, adding thermal disturbance during the melting process to destroy flow stability and to enhance convective heat transfer. The research methods and findings will support the development of energy storage and numerical investigation in other related areas.

physics.flu-dyn↗

arXiv2020

Cracking the Black Box: Distilling Deep Sports Analytics

This paper addresses the trade-off between Accuracy and Transparency for deep learning applied to sports analytics. Neural nets achieve great predictive accuracy through deep learning, and are popular in sports analytics. But it is hard to interpret a neural net model and harder still to extract actionable insights from the knowledge implicit in it. Therefore, we built a simple and transparent model that mimics the output of the original deep learning model and represents the learned knowledge in an explicit interpretable way. Our mimic model is a linear model tree, which combines a collection of linear models with a regression-tree structure. The tree version of a neural network achieves high fidelity, explains itself, and produces insights for expert stakeholders such as athletes and coaches. We propose and compare several scalable model tree learning heuristics to address the computational challenge from datasets with millions of data points.

cs.LG↗

arXiv2018

Toward Interpretable Deep Reinforcement Learning with Linear Model U-Trees

Deep Reinforcement Learning (DRL) has achieved impressive success in many applications. A key component of many DRL models is a neural network representing a Q function, to estimate the expected cumulative reward following a state-action pair. The Q function neural network contains a lot of implicit knowledge about the RL problems, but often remains unexamined and uninterpreted. To our knowledge, this work develops the first mimic learning framework for Q functions in DRL. We introduce Linear Model U-trees (LMUTs) to approximate neural network predictions. An LMUT is learned using a novel on-line algorithm that is well-suited for an active play setting, where the mimic learner observes an ongoing interaction between the neural net and the environment. Empirical evaluation shows that an LMUT mimics a Q function substantially better than five baseline methods. The transparent tree structure of an LMUT facilitates understanding the network's learned knowledge by analyzing feature influence, extracting rules, and highlighting the super-pixels in image inputs.

cs.LG↗

arXiv2018

Deep Reinforcement Learning in Ice Hockey for Context-Aware Player Evaluation

A variety of machine learning models have been proposed to assess the performance of players in professional sports. However, they have only a limited ability to model how player performance depends on the game context. This paper proposes a new approach to capturing game context: we apply Deep Reinforcement Learning (DRL) to learn an action-value Q function from 3M play-by-play events in the National Hockey League (NHL). The neural network representation integrates both continuous context signals and game history, using a possession-based LSTM. The learned Q-function is used to value players' actions under different game contexts. To assess a player's overall performance, we introduce a novel Game Impact Metric (GIM) that aggregates the values of the player's actions. Empirical Evaluation shows GIM is consistent throughout a play season, and correlates highly with standard success measures and future salary.

cs.LG↗