Abstract
The large and diverse availability of mobility data enables the development of predictive models capable of recognizing various types of movements. Through a variety of GPS devices, any moving entity, animal, person, or vehicle can generate spatio-temporal trajectories. This data is used to infer migration patterns, manage traffic in large cities, and monitor the spread and impact of diseases, all critical situations that necessitate a thorough understanding of the underlying problem. Researchers, businesses, and governments use mobility data to make decisions that affect people’s lives in many ways, employing accurate but opaque deep learning models that are difficult to interpret from a human standpoint. To address these limitations, we propose Geolet, a human-interpretable machine-learning model for trajectory classification. We use discriminative sub-trajectories extracted from mobility data to turn trajectories into a simplified representation that can be used as input by any machine learning classifier. We test our approach against state-of-the-art competitors on real-world datasets. Geolet outperforms black-box models in terms of accuracy while being orders of magnitude faster than its interpretable competitors.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Code available at: github.com/cri98li/Geolet.
- 2.
animals: \(\textit{prec}=2\) \(\textit{ns}=21\); vehicles: \(\textit{prec}=6\) \(\textit{ns}=20\); seabirds: \(\textit{prec}=5\) \(\textit{ns}=50\); geolife: \(\textit{prec}=6\) \(\textit{ns}=50\); taxi: \(\textit{prec}=5\) \(\textit{ns}=50\).
- 3.
\(\textit{prec}\in [4, 5, 6, 7]\); \(k\in [2, 5, 20, 100]\); \(w\in [2, 3, 5]\); \(top_{ss}\in [1, 2, 10, 50]\) on the training set. Hyperparameter choice does not significantly affect the method’s performance. We found constant accuracy values for most of the hyperparameters tested. There were, however, peaks in the accuracy score for some values. Thus, for animals we set \(\textit{prec}=4, w=3\text { and }top_{ss}=2\). For the vehicles \(\textit{prec}=6, w=3\text { and }top_{ss}=10\).
- 4.
n_estimators = range(300, 1500, 300), criterion = [gini, entropy], max_depth = range(2, 20, 3).
- 5.
Tests are performed on a machine with CPU: AMD Ryzen 9 3900X; RAM: 32 GB; OS: EndeavourOS Linux. Due to resource limitations, we used 20% of geolife and 70% of taxi.
References
Andrienko, G.L., et al.: (So) big data and the transformation of the city. Int. J. Data Sci. Anal. 11(4), 311–340 (2021)
Bellman, R., Kalaba, R.: On adaptive control processes. IRE Trans. Autom. Control. 4(2), 1–9 (1959)
Bodria, F., Giannotti, F., Guidotti, R., Naretto, F., Pedreschi, D., Rinzivillo, S.: Benchmarking and survey of explanation methods for black box models. CoRR abs/2102.13076 (2021)
Dempster, A., Petitjean, F., Webb, G.I.: ROCKET: exceptionally fast and accurate time series classification using random convolutional kernels. Data Min. Knowl. Discov. 34(5), 1454–1495 (2020)
Ferrero, C.A., Alvares, L.O., Zalewski, W., Bogorny, V.: MOVELETS: exploring relevant subtrajectories for robust trajectory classification. In: SAC, pp. 849–856. ACM (2018)
de Freitas, N.C.A., da Silva, T.L.C., de Macêdo, J.A.F., Junior, L.M.: Using deep learning for trajectory classification in imbalanced dataset. In: FLAIRS Conference (2021)
Guidotti, R., Monreale, A., Ruggieri, S., Turini, F., Giannotti, F., Pedreschi, D.: A survey of methods for explaining black box models. ACM Comput. Surv. (CSUR) 51(5), 1–42 (2018)
Kontopoulos, I., Makris, A., Tserpes, K., Bogorny, V.: Traclets: harnessing the power of computer vision for trajectory classification (2022)
Lee, J., Han, J., Li, X., Gonzalez, H.: TraClass: trajectory classification using hierarchical region-based and trajectory-based clustering. Proc. VLDB Endow. 1(1), 1081–1094 (2008)
Lin, J., Keogh, E.J., Lonardi, S., Chiu, B.Y.: A symbolic representation of time series, with implications for streaming algorithms. In: DMKD, pp. 2–11. ACM (2003)
Miller, T.: Explanation in artificial intelligence: insights from the social sciences. Artif. Intell. 267, 1–38 (2019)
Nguyen, T.L., Ifrim, G.: Mrsqm: fast time series classification with symbolic representations. CoRR abs/2109.01036 (2021). arxiv.org/abs/2109.01036
Petry, L.M., da Silva, C.L., Esuli, A., Renso, C., Bogorny, V.: MARC: a robust method for multiple-aspect trajectory classification via space, time, and semantic embeddings. Int. J. Geogr. Inf. Sci. 34(7), 1428–1450 (2020)
da Silva, C.L., Petry, L.M., Bogorny, V.: A survey and comparison of trajectory classification methods. In: BRACIS, pp. 788–793. IEEE (2019)
Simonyan, K., Vedaldi, A., Zisserman, A.: Deep inside convolutional networks: visualising image classification models and saliency maps (2014)
Suwardi, I.S., Dharma, D., Satya, D.P., Lestari, D.P.: Geohash index based spatial data model for corporate. In: 2015 International Conference on Electrical Engineering and Informatics (ICEEI), pp. 478–483. IEEE (2015)
Tan, P.N., Steinbach, M.S., Kumar, V.: Introduction to Data Mining. Pearson Education India, Noida (2016)
Theissler, A., Spinnato, F., Schlegel, U., Guidotti, R.: Explainable AI for time series classification: a review, taxonomy and research directions. IEEE Access 10, 100700–100724 (2022)
Trasarti, R., Guidotti, R., Monreale, A., Giannotti, F.: Myway: location prediction via mobility profiling. Inf. Syst. 64, 350–367 (2017)
Vouros, A., et al.: A generalised framework for detailed classification of swimming paths inside the morris water maze. Sci. Rep. 8(1), 1–15 (2018)
Xiao, Z., Wang, Y., Fu, K., Wu, F.: Identifying different transportation modes from trajectory data using tree-based ensemble classifiers. ISPRS Int. J. Geo Inf. 6(2), 57 (2017)
Ye, L., Keogh, E.J.: Time series shapelets: a new primitive for data mining. In: KDD, pp. 947–956. ACM (2009)
Acknowledgment
This work is partially supported by the European Union NextGenerationEU programme under the fuding schemes PNRR-PE-AI scheme (M4C2, investment 1.3, line on Artificial Intelligence) FAIR (Future Artificial Intelligence Research), and “SoBigData.it - Strengthening the Italian RI for Social Mining and Big Data Analytics” - Prot. IR0000013. This work is partially supported by the European Community H2020 programme under the funding schemes: H2020-INFRAIA-2019-1: Res. Infr. G.A. 871042 SoBigData++, G.A. 761758 Humane AI, G.A. 952215 TAILOR, ERC-2018-ADG G.A. 834756 XAI, and CHIST-ERA-19-XAI-010 SAI, and by the Green.Dat.AI Horizon Europe research and innovation programme under the G.A. 101070416.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Landi, C., Spinnato, F., Guidotti, R., Monreale, A., Nanni, M. (2023). Geolet: An Interpretable Model for Trajectory Classification. In: Crémilleux, B., Hess, S., Nijssen, S. (eds) Advances in Intelligent Data Analysis XXI. IDA 2023. Lecture Notes in Computer Science, vol 13876. Springer, Cham. https://doi.org/10.1007/978-3-031-30047-9_19
Download citation
DOI: https://doi.org/10.1007/978-3-031-30047-9_19
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-30046-2
Online ISBN: 978-3-031-30047-9
eBook Packages: Computer ScienceComputer Science (R0)