References of "Liu, Feng"
     in
Bookmark and Share    
Full Text
Peer Reviewed
See detailUncertainty quantification in profile Hidden Markov Models (pHMM)-based activity sequences characterization
Saadi, Ismaïl ULg; Liu, Feng; El Saeid Mustafa, Ahmed Mohamed ULg et al

Poster (2016, April)

Recently, Liu et al. (2015) proposed a method to characterize activity sequences stemming from activity-travel diaries. The framework is structured as follows: from an extracted set of activity sequences ... [more ▼]

Recently, Liu et al. (2015) proposed a method to characterize activity sequences stemming from activity-travel diaries. The framework is structured as follows: from an extracted set of activity sequences, (a) the occurrence probabilities of the different activities are determined as well as their sequential order for aligning the activity sequences. Then, (b) profile Hidden Markov Models (pHMM) are defined based on the previous output. This technique is interesting given the fact that it is also able to include the irregular activities and, as a result, their derived trips. In this context, thinking about integration with an agent-based micro-simulation model requires, as a preliminary step, an uncertainty quantification analysis in order to measure the variability of the outcome. This approach is all the more necessary when agent-based micro-simulation is used to predict mid- and long-term system states. [less ▲]

Detailed reference viewed: 50 (5 ULg)
Full Text
Peer Reviewed
See detailA framework to identify housing location patterns using profile Hidden Markov Models
Saadi, Ismaïl ULg; Liu, Feng; El Saeid Mustafa, Ahmed Mohamed ULg et al

in Advanced Science Letters (2016), 22(9), 2117-2121

The determination of comprehensive activity-travel patterns is important in the context of agent-based micro-simulation modelling. This paper presents an improved method based on profile Hidden Markov ... [more ▼]

The determination of comprehensive activity-travel patterns is important in the context of agent-based micro-simulation modelling. This paper presents an improved method based on profile Hidden Markov Models (pHMMs) able to include information related to the agents’ residential locations. As proposed in the framework of Liu et al. (2015), pHMMs only characterize activity-travel patterns from the activity sequences perspective. In this context, information related to the primary activity locations (e.g. home, work) is not available and, as a result, it cannot be extracted from the pHMMs themselves. With respect to this limitation, we propose to apply the framework of Liu et al. (2015) with an extension to include characterization of residential locations. Following the established guidelines, the activity sequences and their related residential locations are extracted from the activity-travel diaries in order to estimate the regularity of the activities as well as their sequential order. Subsequently, within each residential activity, we include a categorization at an aggregate level (provinces). The methodology is powerful as it characterizes any length of sequence, allowing the generation of unlimited agent plans with information about residential location. Regarding data collection, the activity-travel diaries are provided by the Belgian Household Daily Travel Survey (2010). The results obtained after the simulations indicate a good match between the predicted and observed residential locations at both the national and provincial levels. [less ▲]

Detailed reference viewed: 85 (29 ULg)
Full Text
Peer Reviewed
See detailIdentifying mismatch between urban travel demand and transport network services using GPS data: A case study in the fast growing Chinese city of Harbin
Cui, JianXun; Liu, Feng; Hu, Jia et al

in Neurocomputing (2016), 181

The world's urban population growth and economic development have led to the reshaping of metropolitan space layouts among residential, employment and shopping locations, generating growing mismatch ... [more ▼]

The world's urban population growth and economic development have led to the reshaping of metropolitan space layouts among residential, employment and shopping locations, generating growing mismatch between travel demand and transport services. A reliable method to accurately analyze mobility demand and underlying transport network systems and to identify areas with serious mismatch problems is important for the design of effective policy measures. In this paper, we make use of the wide deployment of GPS devices in vehicles in many cities today, to develop such a method. This approach is developed using GPS data collected from all taxis operating in the Chinese city of Harbin between July and September in 2013. It consists of four major steps. First, city-wide mobility patterns are modeled based on GPS trajectories. This model captures a set of key traffic characteristics for each pair of regions in the entire urban network, including travel demand, travel speed and route directness of travel paths. From this model, a set of indicators is then built to measure the road transport performance between the regions, and the areas with serious mismatch problems are subsequently pinpointed. Finally, the identified problematic regions are further examined and specific transport problems are analyzed. By applying the proposed method to the city of Harbin, the potential and effectiveness of this method are demonstrated. Moreover, with more and more urban vehicles being equipped with GPS devices, the designed method can be easily transferred to other cities, thus paving a way for the adoption of the presented approach for an up-to-date and spatial-temporal sensitive road network analysis approach that supports the establishment of a more sustainable urban transport system. [less ▲]

Detailed reference viewed: 35 (3 ULg)
Full Text
Peer Reviewed
See detailDetecting urban road network accessibility problems using taxi GPS data
Cui, JianXun; Liu, Feng; Janssens, Davy et al

in Journal of Transport Geography (2016), 51

Urban population growth and economic development have led to the creation of new communities, jobs and services at places where the existing road network might not cover or efficiently handle traffic ... [more ▼]

Urban population growth and economic development have led to the creation of new communities, jobs and services at places where the existing road network might not cover or efficiently handle traffic. This generates isolated pockets of areas which are difficult to reach through the transport system. To address this accessibility problem, we have developed a novel approach to systematically examine the current urban land use and road network conditions as well as to identify poorly connected regions, using GPS data collected from taxis. This method is composed of four major steps. First, city-wide passenger travel demand patterns and travel times are modeled based on GPS trajectories. Upon this model, high density residential regions are then identified, and measures to assess accessibility of each of these places are developed. Next, the regions with the lowest level of accessibility among all the residential areas are detected, and finally the detected regions are further examined and specific transport situations are analyzed. By applying the proposed method to the Chinese city of Harbin, we have identified 20 regions that have the lowest level of accessibility by car among all the identified residential areas. A serious reachability problem to petrol stations has also been discovered, in which drivers from 92.6% of the residential areas have to travel longer than 30 min to refill their cars. Furthermore, the comparison against a baseline model reveals the capacity of the derived measures in accounting for the actual travel routes under divergent traffic conditions. The experimental results demonstrate the potential and effectiveness of the proposed method in detecting car-based accessibility problems, contributing towards the development of urban road networks into a system that has better reachability and more reduced inequity. [less ▲]

Detailed reference viewed: 29 (1 ULg)
Full Text
Peer Reviewed
See detailCharacterizing activity sequences using Profile Hidden Markov Models
Liu, Feng; Janssens, Davy; Cui, JianXun et al

in Expert Systems with Applications (2015), 42(13), 57055722

In literature, activity sequences, generated from activity-travel diaries, have been analyzed and classified into clusters based on the composition and ordering of the activities using Sequence Alignment ... [more ▼]

In literature, activity sequences, generated from activity-travel diaries, have been analyzed and classified into clusters based on the composition and ordering of the activities using Sequence Alignment Methods (SAM). However, using these methods, only the frequent activities in each cluster are extracted and qualitatively described; the infrequent activities and their related travel episodes are disregarded. Thus, to quantify the occurrence probabilities of all the daily activities as well as their sequential orders, we develop a novel process to build multiple alignments of the sequences and subsequently derive profile Hidden Markov Models (pHMMs). This process consists of 4 major steps. First, activity sequences are clustered based on a pre-defined scheme. The frequent activities along with their sequential orders are then identified in each cluster, and they are subsequently used as a template to guide the construction of a multiple alignment of the cluster of sequences. Finally, a pHMM is employed to convert the multiple alignment into a position-specific scoring system, representing the probability of each frequent activity at each important position of the alignment as well as the probabilities of both insertion and deletion of infrequent activities. By applying the derived pHMMs to a set of activity-travel diaries collected in Belgium as well as a group of mobile phone call location data recorded in Switzerland, the potential and effectiveness of the models in capturing the sequential features of each cluster and distinguishing them from those of other clusters, are demonstrated. The proposed method can also be utilized to improve activity-based transportation model validation and travel survey designs. Furthermore, it offers a wide application in characterizing a group of any related sequences, particularly sequences varying in length and with a high frequency of short sequences that are typically present in human behavior. [less ▲]

Detailed reference viewed: 50 (6 ULg)
Full Text
Peer Reviewed
See detailBuilding a validation measure for activity-based transportation models based on mobile phone data
Liu, Feng; Janssens, Davy; Cui, JianXun et al

in Expert Systems with Applications (2014), 41(14), 6174-6189

Activity-based micro-simulation transportation models typically predict 24-h activity-travel sequences for each individual in a study area. These sequences serve as a key input for travel demand analysis ... [more ▼]

Activity-based micro-simulation transportation models typically predict 24-h activity-travel sequences for each individual in a study area. These sequences serve as a key input for travel demand analysis and forecasting in the region. However, despite their importance, the lack of a reliable benchmark to evaluate the generated sequences has hampered further development and application of the models. With the wide deployment of mobile phone devices today, we explore the possibility of using the travel behavioral information derived from mobile phone data to build such a validation measure. Our investigation consists of three steps. First, the daily trajectory of locations, where a user performed activities, is constructed from the mobile phone records. To account for the discrepancy between the stops revealed by the call data and the real location traces that the user has made, the daily trajectories are then transformed into actual travel sequences. Finally, all the derived sequences are classified into typical activity-travel patterns which, in combination with their relative frequencies, define an activity-travel profile. The established profile characterizes the current activity-travel behavior in the study area, and can thus be used as a benchmark for the assessment of the activity-based transportation models. By comparing the activity-travel profiles derived from the call data with statistics that stem from traditional activity-travel surveys, the validation potential is demonstrated. In addition, a sensitivity analysis is carried out to assess how the results are affected by the different parameter settings defined in the profiling process. [less ▲]

Detailed reference viewed: 65 (5 ULg)
Full Text
Peer Reviewed
See detailSemantic Annotation of Global Positioning System Traces: Activity Type Inference
Reumers, Sofie; Liu, Feng; Janssens, Davy et al

in Transportation Research Record: Journal of the Transportation Research Board (2013), 2383

Because of the rapid development of technology, larger data sets on activity travel behavior have become available. These data sets often lack semantic interpretation. This lack of interpretation implies ... [more ▼]

Because of the rapid development of technology, larger data sets on activity travel behavior have become available. These data sets often lack semantic interpretation. This lack of interpretation implies that annotation of activity type and transportation mode is necessary. This paper aims to infer activity types from Global Positioning System (GPS) traces by developing a decision tree-based model. The model considers only activity start times and activity durations. On the basis of the decision tree classification, a probability distribution and a point prediction model were constructed. The probability matrix described the probability of each activity type for each class (i.e., combination of activity start time and activity duration). In each class, the point prediction model selected the activity type that had the highest probability. Two types of data were collected in 2006 and 2007 in Flanders, Belgium (i.e., activity travel data and GPS data). The optimal classification tree constructed contained 18 leaves. Consequently, 18 if-then rules were derived. An accuracy of 74% was achieved when the tree was trained. The accuracy of the model for the validation set (72.5%) showed that overfitting was minimal. When the model was applied to the test set, the accuracy was almost 76%. The models indicated the importance of time information in the semantic enrichment process. This study contributes to future data collection in that it enables researchers to infer activity types directly from activity start time and duration information obtained from GPS data. Because no location information is needed, this research can be easily and readily applied to millions of individual agents. [less ▲]

Detailed reference viewed: 158 (5 ULg)
Full Text
Peer Reviewed
See detailProfiling workers’ activity-travel behavior based on mobile phone data
Liu, Feng; Janssens, Davy; Wets, Geert et al

in Proceedings of the Third International Conference on the Analysis of Mobile Phone Datasets (NetMob) (2013)

Activity-based micro-simulation models typically predict 24-hour activity-travel patterns for each individual in a study area. These patterns reflect the characteristics of the available transportation ... [more ▼]

Activity-based micro-simulation models typically predict 24-hour activity-travel patterns for each individual in a study area. These patterns reflect the characteristics of the available transportation infrastructure and land-use system as well as individuals’ lifestyles and needs. However, the lack of a reliable benchmark to evaluate the generated patterns has been a major concern. To address this issue, we explore the possibility of using mobile phone data to build such a validation measure. Our investigation consists of three steps. First, the daily trajectory of locations, where a user performed activities, is constructed from the mobile phone records. To account for the discrepancy between the movements revealed by the call data and the real traces that the user has made, the daily trajectories are then transformed into travel sequences. Finally, all the inferred travel sequences are classified into typical activity-travel patterns which, in combination with their relative frequencies, define a profile. The established profile represents the activity-travel behavior in the study area, and thus can be used as a benchmark for the validation of the activity-based models. By comparing the benchmark profiles derived from the call data with statistics that stem from activity-travel surveys, the validation potential is demonstrated. In addition, a sensitivity analysis is carried out to assess how the results are affected by the different parameter settings defined in the profiling process. [less ▲]

Detailed reference viewed: 92 (0 ULg)
Full Text
Peer Reviewed
See detailAnnotating mobile phone location data with activity purposes using machine learning algorithms
Liu, Feng; Janssens, Davy; Wets, Geert et al

in Expert Systems with Applications (2013), 40(8), 32993311

Individual human travel patterns captured by mobile phone data have been quantitatively characterized by mathematical models, but the underlying activities which initiate the movement are still in a less ... [more ▼]

Individual human travel patterns captured by mobile phone data have been quantitatively characterized by mathematical models, but the underlying activities which initiate the movement are still in a less-explored stage. As a result of the nature of how activity and related travel decisions are made in daily life, human activity-travel behavior exhibits a high degree of spatial and temporal regularities as well as sequential ordering. In this study, we investigate to what extent the behavioral routines could reveal the activities being performed at mobile phone call locations that are captured when users initiate or receive a voice call or message. Our exploration consists of four steps. First, we define a set of comprehensive temporal variables characterizing each call location. Feature selection techniques are then applied to choose the most effective variables in the second step. Next, a set of state-of-the-art machine learning algorithms including Support Vector Machines, Logistic Regression, Decision Trees and Random Forests are employed to build classification models. Alongside, an ensemble of the results of the above models is also tested. Finally, the inference performance is further enhanced by a post-processing algorithm. Using data collected from natural mobile phone communication patterns of 80 users over a period of more than one year, we evaluated our approach via a set of extensive experiments. Based on the ensemble of the models, we achieved prediction accuracy of 69.7%. Furthermore, using the post processing algorithm, the performance obtained a 7.6% improvement. The experiment results demonstrate the potential to annotate mobile phone locations based on the integration of data mining techniques with the characteristics of underlying activity-travel behavior, contributing towards the semantic comprehension and further application of the massive data. [less ▲]

Detailed reference viewed: 128 (2 ULg)
Full Text
Peer Reviewed
See detailSemantic Annotation of GPS Traces: Activity Type Inference
Reumers, Sofie; Liu, Feng; Janssens, Davy et al

in Proceedings of the 92nd Annual Meeting of the Transportation Research Board (DVD-ROM) (2013)

Due to the rapid development of technology, larger data sets concerning activity travel behavior become available. These data sets often lack semantic interpretation. This implies that annotation in terms ... [more ▼]

Due to the rapid development of technology, larger data sets concerning activity travel behavior become available. These data sets often lack semantic interpretation. This implies that annotation in terms of activity type and transportation mode is necessary. This paper aims to infer activity types from GPS traces by developing a decision tree-based model. The model only considers activity start times and activity durations. Based on the decision tree classification, a probability distribution and a point prediction model were constructed. The probability matrix describes the probability of each activity type for each class (i.e. combination of activity start time and activity duration). In each class, the point prediction model selects the activity type that has the highest probability. Two types of data were collected in 2006 and 2007 in Flanders, Belgium, i.e. activity travel data and GPS data. The optimal classification tree constructed comprises 18 leaves. Consequently, 18 if-then rules were derived. An accuracy of 74% was achieved when training the tree. The accuracy of the model for the validation set, i.e. 72.5%, shows that overfitting is minimal. When applying the model to the test set, the accuracy was almost 76%. The models indicate the importance of time information in the semantic enrichment process. This study contributes to future data collection in that it enables researchers to directly infer activity types from activity start time and duration information obtained from GPS data. Because no location information is needed, this research can be easily and readily implemented to millions of individual agents. [less ▲]

Detailed reference viewed: 156 (0 ULg)