\n \n \n
\n
\n\n \n \n \n \n \n \n New Complexity Results on Aggregating Lexicographic Preferences Trees Using Positional Scoring Rules.\n \n \n \n \n\n\n \n Liu, X.; and Truszczynski, M.\n\n\n \n\n\n\n In
Proceedings of the 6th International Conference on Algorithmic Decision Theory (ADT), 2019. Springer\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 13 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@inproceedings{conf/adt19/LiuT,\n author = {Xudong Liu and Miroslaw Truszczynski},\n booktitle = {Proceedings of the 6th International Conference on Algorithmic Decision Theory (ADT)},\n publisher = {Springer},\n\turl="https://link.springer.com/chapter/10.1007/978-3-030-31489-7_7",\n title = {New Complexity Results on Aggregating Lexicographic Preferences Trees Using Positional Scoring Rules},\n abstract = {Aggregating votes that are preference orders over candidates or alternatives is a fundamental problem of decision theory and social choice. We study this problem in the setting when alternatives are described as tuples of values of attributes. The combinatorial spaces of such alternatives make explicit enumerations of alternatives from the most to the least preferred infeasible. Instead, votes may be specified implicitly in terms of some compact and intuitive preference representation mechanism. In our work, we assume that votes are given as lexicographic preference trees and consider two preference-aggregation problems, the winner problem and the evaluation problem. We study them under the assumption that positional scoring rules are used for aggregation. In particular, we consider k-Approval and b-Borda, a generalized Borda rule, and we discover new computational complexity results for them.},\n year = 2019\n}\n\n
\n
\n\n\n
\n Aggregating votes that are preference orders over candidates or alternatives is a fundamental problem of decision theory and social choice. We study this problem in the setting when alternatives are described as tuples of values of attributes. The combinatorial spaces of such alternatives make explicit enumerations of alternatives from the most to the least preferred infeasible. Instead, votes may be specified implicitly in terms of some compact and intuitive preference representation mechanism. In our work, we assume that votes are given as lexicographic preference trees and consider two preference-aggregation problems, the winner problem and the evaluation problem. We study them under the assumption that positional scoring rules are used for aggregation. In particular, we consider k-Approval and b-Borda, a generalized Borda rule, and we discover new computational complexity results for them.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Voting-based Ensemble Learning for Partial Lexicographic Preference Forests over Combinatorial Domains.\n \n \n \n \n\n\n \n Liu, X.; and Truszczynski, M.\n\n\n \n\n\n\n
Annals of Mathematics and Artificial Intelligence, Springer, 87: 137-155. 2019.\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 16 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{j/AMAI19/LiuT,\n author = {Xudong Liu and Miroslaw Truszczynski},\n title = {Voting-based Ensemble Learning for Partial Lexicographic Preference Forests over Combinatorial Domains},\n journal = {Annals of Mathematics and Artificial Intelligence, Springer},\n volume = {87}, \n pages = {137-155},\n issue = {1-2},\n publisher = {Springer},\n abstract = {We study preference representation models based on partial lexicographic preference trees\n(PLP-trees). We propose to represent preference relations as forests of small PLP-trees\n(PLP-forests), and to use voting rules to aggregate orders represented by the individual\ntrees into a single order to be taken as a model of the agent’s preference relation. We show\nthat when learned from examples, PLP-forests have better accuracy than single PLP-trees.\nWe also show that the choice of a voting rule does not have a major effect on the aggregated order, thus \nrendering the problem of selecting the “right” rule less critical. Next, for\nthe proposed PLP-forest preference models, we develop methods to compute optimal and\nnear-optimal outcomes, the tasks that appear difficult for some other common preference\nmodels. Lastly, we compare our models with those based on decision trees, which brings up\nquestions for future research.},\n\tissn="1573-7470", \n\tdoi="10.1007/s10472-019-09645-7", \n\turl="https://doi.org/10.1007/s10472-019-09645-7",\n year = 2019\n}\n\n
\n
\n\n\n
\n We study preference representation models based on partial lexicographic preference trees (PLP-trees). We propose to represent preference relations as forests of small PLP-trees (PLP-forests), and to use voting rules to aggregate orders represented by the individual trees into a single order to be taken as a model of the agent’s preference relation. We show that when learned from examples, PLP-forests have better accuracy than single PLP-trees. We also show that the choice of a voting rule does not have a major effect on the aggregated order, thus rendering the problem of selecting the “right” rule less critical. Next, for the proposed PLP-forest preference models, we develop methods to compute optimal and near-optimal outcomes, the tasks that appear difficult for some other common preference models. Lastly, we compare our models with those based on decision trees, which brings up questions for future research.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Hourly Weather Data Projection due to Climate Change for Impact Assessment on Building and Infrastructure.\n \n \n \n \n\n\n \n Jiang, A.; Liu, X.; Czarnecki, E.; and Zhang, C.\n\n\n \n\n\n\n
Sustainable Cities and Society, Elsevier. 2019.\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 5 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{j/SCS19/JLiuCZ,\n author = {Aiyin Jiang and Xudong Liu and Emily Czarnecki and Chengyi Zhang},\n title = {Hourly Weather Data Projection due to Climate Change for Impact Assessment on Building and Infrastructure},\n journal = {Sustainable Cities and Society, Elsevier},\n publisher = {Elsevier},\n\tissn = "2210-6707",\n\tdoi = "https://doi.org/10.1016/j.scs.2019.101688",\n\turl = "http://www.sciencedirect.com/science/article/pii/S2210670719304810",\n\tabstract = {The global climate change research has been conducted for a few years in various professional communities. \nIn the building industry, researchers usually investigate the future building energy demands due to the climate change \nby simulation software. The input les to the simulation software includes projected weather data and building models. \nAlthough there exist a few mathematical methods to project the future weather, the morphing method is the most well-known among them. \nIn the meantime, the simulation software and weather data are in a variety of formats depending on country of origin and/or simulation package. \nIn order to provide both the research and the professional communities the possibility to undertake climate change impact assessments on buildings, \ncoastal engineering and construction, land use and other related areas, this study develops the web-based application Weather Morph: \nClimate Change Weather File Generator accessible to generate the future weather data for more than 2100 locations throughout the world for all \nfour IPCC (Intergovernmental Panel of Climate Change) emission scenarios in the three future time slices of the 2020s, 2050s and 2080s. \nThe output of the application is projected future weather datasets in formats TMY2 and EPW for general use.},\n year = "2019"\n}\n\n
\n
\n\n\n
\n The global climate change research has been conducted for a few years in various professional communities. In the building industry, researchers usually investigate the future building energy demands due to the climate change by simulation software. The input les to the simulation software includes projected weather data and building models. Although there exist a few mathematical methods to project the future weather, the morphing method is the most well-known among them. In the meantime, the simulation software and weather data are in a variety of formats depending on country of origin and/or simulation package. In order to provide both the research and the professional communities the possibility to undertake climate change impact assessments on buildings, coastal engineering and construction, land use and other related areas, this study develops the web-based application Weather Morph: Climate Change Weather File Generator accessible to generate the future weather data for more than 2100 locations throughout the world for all four IPCC (Intergovernmental Panel of Climate Change) emission scenarios in the three future time slices of the 2020s, 2050s and 2080s. The output of the application is projected future weather datasets in formats TMY2 and EPW for general use.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Human-In-The-Loop Learning of Qualitative Preference Models.\n \n \n \n \n\n\n \n Allen, J.; Moussa, A.; and Liu, X.\n\n\n \n\n\n\n In
Proceedings of the 32nd International Florida Artificial Intelligence Research Society Conference (FLAIRS), pages 108-111, 2019. AAAI Press\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 5 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@inproceedings{conf/flairs32/AMLiu,\n author = {Joseph Allen and Ahmed Moussa and Xudong Liu},\n booktitle = {Proceedings of the 32nd International Florida Artificial Intelligence Research Society Conference (FLAIRS)},\n title = {Human-In-The-Loop Learning of Qualitative Preference Models},\n pages = {108-111},\n publisher = {AAAI Press},\n\turl="https://aaai.org/ocs/index.php/FLAIRS/FLAIRS19/paper/view/18284",\n\tabstract = {In this work, we present a novel human-in-the-loop framework to help the human user understand the decision \nmaking process that involves choosing preferred options. We focus on qualitative preference models over alternatives from \ncombinatorial domains. This framework is interactive: the user provides her behavioral data to the framework, and the \nframework explains the learned model to the user. It is iterative: the framework collects feedback on the learned model from \nthe user and tries to improve it accordingly till the user terminates the iteration. In order to communicate the \nlearned preference model to the user, we develop visualization of intuitive and explainable graphic models, such as lexicographic \npreference trees and forests, and conditional preference networks. \nTo this end, we discuss key aspects of our framework for lexicographic preference models.\n},\n year = 2019\n}\n\n
\n
\n\n\n
\n In this work, we present a novel human-in-the-loop framework to help the human user understand the decision making process that involves choosing preferred options. We focus on qualitative preference models over alternatives from combinatorial domains. This framework is interactive: the user provides her behavioral data to the framework, and the framework explains the learned model to the user. It is iterative: the framework collects feedback on the learned model from the user and tries to improve it accordingly till the user terminates the iteration. In order to communicate the learned preference model to the user, we develop visualization of intuitive and explainable graphic models, such as lexicographic preference trees and forests, and conditional preference networks. To this end, we discuss key aspects of our framework for lexicographic preference models. \n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Learning Optimal and Near-Optimal Lexicographic Preference Lists.\n \n \n \n \n\n\n \n Moussa, A.; and Liu, X.\n\n\n \n\n\n\n In
Proceedings of the 32nd International Florida Artificial Intelligence Research Society Conference (FLAIRS), pages 128-131, 2019. AAAI Press\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 3 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@inproceedings{conf/flairs32/MLiu,\n author = {Ahmed Moussa and Xudong Liu},\n booktitle = {Proceedings of the 32nd International Florida Artificial Intelligence Research Society Conference (FLAIRS)},\n title = {Learning Optimal and Near-Optimal Lexicographic Preference Lists},\n pages = {128-131},\n publisher = {AAAI Press},\n\turl="https://aaai.org/ocs/index.php/FLAIRS/FLAIRS19/paper/view/18289",\n\tabstract = {We consider learning problems of an intuitive and concise preference model, called lexicographic preference lists (LP-lists). \nGiven a set of examples that are pairwise ordinal preferences over a universe of objects built of attributes of discrete values, \nwe want to learn (1) an optimal LP-list that decides the maximum number of these examples, or (2) a near-optimal LP-list that decides as many examples as it \ncan. To this end, we introduce a dynamic programming based algorithm and a genetic algorithm for these two learning problems, respectively. \nFurthermore, we empirically demonstrate that the sub-optimal models computed by the genetic algorithm very well approximate the de facto \noptimal models computed by our dynamic programming based algorithm, and that the genetic algorithm outperforms the baseline greedy heuristic with \nhigher accuracy predicting new preferences.},\n year = 2019\n}\n\n
\n
\n\n\n
\n We consider learning problems of an intuitive and concise preference model, called lexicographic preference lists (LP-lists). Given a set of examples that are pairwise ordinal preferences over a universe of objects built of attributes of discrete values, we want to learn (1) an optimal LP-list that decides the maximum number of these examples, or (2) a near-optimal LP-list that decides as many examples as it can. To this end, we introduce a dynamic programming based algorithm and a genetic algorithm for these two learning problems, respectively. Furthermore, we empirically demonstrate that the sub-optimal models computed by the genetic algorithm very well approximate the de facto optimal models computed by our dynamic programming based algorithm, and that the genetic algorithm outperforms the baseline greedy heuristic with higher accuracy predicting new preferences.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n An Extensible and Personalizable Multi-Modal Trip Planner.\n \n \n \n \n\n\n \n Liu, X.; Fritz, C.; and Klenk, M.\n\n\n \n\n\n\n In
Proceedings of the 32nd International Florida Artificial Intelligence Research Society Conference (FLAIRS), pages 124-127, 2019. AAAI Press\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 164 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@inproceedings{conf/flairs32/LiuFK,\n author = {Xudong Liu and Christian Fritz and Matthew Klenk},\n booktitle = {Proceedings of the 32nd International Florida Artificial Intelligence Research Society Conference (FLAIRS)},\n title = {An Extensible and Personalizable Multi-Modal Trip Planner},\n pages = {124-127},\n publisher = {AAAI Press},\n\turl="https://aaai.org/ocs/index.php/FLAIRS/FLAIRS19/paper/view/18288",\n\tabstract = {Despite a tremendous amount of work in the literature and in the commercial sectors, current approaches to multi-modal \ntrip planning still fail to consistently generate plans that users deem optimal in practice. We believe that this is \ndue to the fact that current planners fail to capture the true preferences of users, e.g., their preferences depend on aspects that are not modeled. \nAn example of this could be a preference not to walk through an unsafe area at night. We present a novel multi-modal trip planner that allows users to \nupload auxiliary geographic data (e.g., crime rates) and to specify temporal constraints and preferences over these data in combination with typical metrics \nsuch as time and cost. Concretely, our planner supports the modes walking, biking, driving, public transit, and taxi, uses linear temporal logic to \ncapture temporal constraints, and preferential cost functions to represent preferences. We show by examples that this allows the expression \nof very interesting preferences and constraints that, naturally, lead to quite diverse optimal plans.},\n year = 2019\n}\n\n
\n
\n\n\n
\n Despite a tremendous amount of work in the literature and in the commercial sectors, current approaches to multi-modal trip planning still fail to consistently generate plans that users deem optimal in practice. We believe that this is due to the fact that current planners fail to capture the true preferences of users, e.g., their preferences depend on aspects that are not modeled. An example of this could be a preference not to walk through an unsafe area at night. We present a novel multi-modal trip planner that allows users to upload auxiliary geographic data (e.g., crime rates) and to specify temporal constraints and preferences over these data in combination with typical metrics such as time and cost. Concretely, our planner supports the modes walking, biking, driving, public transit, and taxi, uses linear temporal logic to capture temporal constraints, and preferential cost functions to represent preferences. We show by examples that this allows the expression of very interesting preferences and constraints that, naturally, lead to quite diverse optimal plans.\n
\n\n\n
\n\n\n\n\n\n