Jorge Freire de Sousa

Ensemble approaches for regression: A survey

The goal of ensemble regression is to combine several models in order to improve the prediction accuracy in learning problems with a numerical target variable. The process of ensemble learning can be divided into three phases: the generation phase, the pruning phase, and the integration phase. We discuss different approaches to each of these phases that are able to deal with the regression problem, categorizing them in terms of their relevant characteristics and linking them to contributions from different fields. Furthermore, this work makes it possible to identify interesting areas for future research.

Improving Mass Transit Operations by Using AVL-Based Systems: A Survey

Intelligent transportation systems based on automated data collection frameworks are widely used by the major transit companies around the globe. This paper describes the current state of the art on improving both planning and control on public road transportation companies using automatic vehicle location (AVL) data. By surveying this topic, the expectation is to help develop a better understanding of the nature, approaches, challenges, and opportunities with regard to these problems. This paper starts by presenting a brief review on improving the network definition based on historical location-based data. Second, it presents a comprehensive review on AVL-based evaluation techniques of the schedule plan (SP) reliability, discussing the existing metrics. Then, the different dimensions on improving the SP reliability are presented in detail, as well as the works addressing such problem. Finally, the automatic control strategies are also revised, along with the research employed over the location-based data. A comprehensive discussion on the techniques employed is provided to encourage those who are starting research on this topic. It is important to highlight that there are still gaps in AVL-based literature, such as the following: 1) long-term travel time prediction; 2) finding optimal slack time; or 3) choosing the best control strategy to apply in each situation in the event of schedule instability. Hence, this paper includes introductory model formulations, reference surveys, formal definitions, and an overview of a promising area, which is of interest to any researcher, regardless of the level of expertise.

Comparing state-of-the-art regression methods for long term travel time prediction

Long-term travel time prediction TTP can be an important planning tool for both freight transport and public transport companies. In both cases it is expected that the use of long-term TTP can improve the quality of the planned services by reducing the error between the actual and the planned travel times. However, for reasons that we try to stretch out along this paper, long-term TTP is almost not mentioned in the scientific literature. In this paper we discuss the relevance of this study and compare three non-parametric state-of-the-art regression methods: Projection Pursuit Regression PPR, Support Vector Machine SVM and Random Forests RF. For each one of these methods we study the best combination of input parameters. We also study the impact of different methods for the pre-processing tasks feature selection, example selection and domain values definition in the accuracy of those algorithms. We use bus travel times data from a bus dispatch system. From an off-the-shelf point-of-view, our experiments show that RF is the most promising approach from the three we have tested. However, it is possible to obtain more accurate results using PPR but with extra pre-processing work, namely on example selection and domain values definition.

An Incremental Probabilistic Model to Predict Bus Bunching in Real-Time

In this paper, we presented a probabilistic framework to predict Bus Bunching (BB) occurrences in real-time. It uses both historical and real-time data to approximate the headway distributions on the further stops of a given route by employing both offline and online supervised learning techniques. Such approximations are incrementally calculated by reusing the latest prediction residuals to update the further ones. These update rules extend the Perceptron’s delta rule by assuming an adaptive beta value based on the current context. These distributions are then used to compute the likelihood of forming a bus platoon on a further stop - which may trigger an threshold-based BB alarm. This framework was evaluated using real-world data about the trips of 3 bus lines throughout an year running on the city of Porto, Portugal. The results are promising.

Setting the length of the planning horizon in the vehicle replacement problem

In some formulations of the vehicle replacement problem, in particular those leading to repair limit type models, the alternative policies are evaluated and compared over a fixed planning horizon. Although it has been widely recognised that the optimal policies derived under these formulations depend critically on the length of the horizon, no method has been presented so far to set appropriately this parameter. In this paper, the authors describe a method which overcomes this shortcoming. Once the best policy has been derived from a given finite horizon with length H, such a policy is repeated indefinitely over time and an equivalent annual rent is computed. The parametrisation of H leads to the definition of an annual rent function with a sequence of nearly equidistant local minima. It is suggested that in practice the second local minimum of this function leads to an adequate choice of the parameter H. The method can be applied both to stochastic and deterministic cost modelling situations. The method was tested using both real data from large samples of different types of passenger vehicles and artificially generated data.

Validation of both number and coverage of bus schedules using AVL data

It is well known that the definition of bus schedules is critical for the service reliability of public transports. Several proposals have been suggested, using data from Automatic Vehicle Location (AVL) systems, in order to enhance the reliability of public transports. In this paper we study the optimum number of schedules and the days covered by each one of them, in order to increase reliability. We use the Dynamic Time Warping distance in order to calculate the similarities between two different dimensioned irregularly spaced data sequences before the use of data clustering techniques. The application of this methodology with the K-Means for a specific bus route demonstrated that a new schedule for the weekends in non-scholar periods could be considered due to its distinct profile from the remaining days. For future work, we propose to apply this methodology to larger data sets in time and in number, corresponding to different bus routes, in order to find a consensual cluster between all the routes.

Improving the accuracy of long-term travel time prediction using heterogeneous ensembles

This paper is about long-term travel time prediction in public transportation. However, it can be useful for a wider area of applications. It follows a heterogeneous ensemble approach with dynamic selection. A vast set of experiments with a pool of 128 tuples of algorithms and parameter sets (a&ps) has been conducted for each of the six studied routes. Three different algorithms, namely, random forest, projection pursuit regression and support vector machines, were used. Then, ensembles of different sizes were obtained after a pruning step. The best approach to combine the outputs is also addressed. Finally, the best ensemble approach for each of the six routes is compared with the best individual a&ps. The results confirm that heterogeneous ensembles are adequate for long-term travel time prediction. Namely, they achieve both higher accuracy and robustness along time than state-of-the-art learners.

Ensemble Learning: A Study on Different Variants of the Dynamic Selection Approach

Integration methods for ensemble learning can use two different approaches: combination or selection. The combination approach (also called fusion) consists on the combination of the predictions obtained by different models in the ensemble to obtain the final ensemble prediction. The selection approach selects one (or more) models from the ensemble according to the prediction performance of these models on similar data from the validation set. Usually, the method to select similar data is the k-nearest neighbors with the Euclidean distance. In this paper we discuss other approaches to obtain similar data for the regression problem. We show that using similarity measures according to the target values improves results. We also show that selecting dynamically several models for the prediction task increases prediction accuracy comparing to the selection of just one model.

Improving SVM-Linear predictions using CART for example selection

This paper describes the study on example selection in regression problems using μ-SVM (Support Vector Machine) linear as prediction algorithm. The motivation case is a study done on real data for a problem of bus trip time prediction. In this study we use three different training sets: all the examples, examples from past days similar to the day where prediction is needed, and examples selected by a CART regression tree. Then, we verify if the CART based example selection approach is appropriate on different regression data sets. The experimental results obtained are promising.

A Multi-Attribute Ranking Solutions Confirmation Procedure

Ranking problems arise from the knowledge of several binary relations defined on a set of alternatives, which we intend to rank. In a previous work, the authors introduced a tool to confirm the solutions of multi-attribute ranking problems as linear extensions of a weighted sum of preference relations. An extension of this technique allows the recognition of critical preference pairs of alternatives, which are often caused by inconsistencies. Herein, a confirmation procedure is introduced and applied to confirm the results obtained by a multi-attribute decision methodology on a tender for the supply of buses to the Porto Public Transport Operator.