Preserving Geo-Indistinguishability of the Emergency Scene to Predict Ambulance Response Time

Arcolezi, Héber H.; Cerna, Selene; Guyeux, Christophe; Couchot, Jean-François

doi:10.3390/mca26030056

Cited by 8 publications

(11 citation statements)

References 39 publications

(102 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, even ML models trained with raw data can also indirectly reveal sensitive information [17,50,16,49], in particular, RNNs [58]. To protect ML models against such threats, under the state-ofthe-art DP guarantee [22,23], there exist some privacypreserving ML alternatives adopted in the literature, e.g., input [19,31,24,29,10,9], gradient [4,33,51,60,48], and objective perturbation [18].…”

Section: Discussion and Related Workmentioning

confidence: 99%

Differentially private multivariate time series forecasting of aggregated human mobility with deep learning: Input or gradient perturbation?

Arcolezi

Couchot

Renaud

et al. 2022

Neural Comput & Applic

Self Cite

View full text Add to dashboard Cite

This paper investigates the problem of forecasting multivariate aggregated human mobility while preserving the privacy of the individuals concerned. Differential privacy, a state-of-the-art formal notion, has been used as the privacy guarantee in two different and independent steps when training deep learning models. On one hand, we considered gradient perturbation, which uses the differentially private stochastic gradient descent algorithm to guarantee the privacy of each time series sample in the learning stage. On the other hand, we considered input perturbation, which adds differential privacy guarantees in each sample of the series before applying any learning. We compared four state-of-the-art recurrent neural networks: Long Short-Term Memory, Gated Recurrent Unit, and their Bidirectional architectures, i.e., Bidirectional-LSTM and Bidirectional-GRU. Extensive experiments were conducted with a real-world multivariate mobility dataset, which we published openly along with this paper. As shown in the results, differentially private deep learning models trained under gradient or input perturbation achieve nearly the same performance as non-private deep learning models, with loss in performance varying between 0.57% to 2.8%. The contribution of this Héber H.

show abstract

Section: Discussion and Related Workmentioning

confidence: 99%

Differentially private multivariate time series forecasting of aggregated human mobility with deep learning: Input or gradient perturbation?

Arcolezi

Couchot

Renaud

et al. 2022

Neural Comput & Applic

Self Cite

View full text Add to dashboard Cite

show abstract

“…• LGBM: objective="binary", is unbalance=True, importance type="gain", boosting type="gbdt", and the search space was n estimators [50-1000], learning rate [0.001-0.8], max depth [1][2][3][4][5][6][7][8][9][10], and colsample by tree [0.5-1]. • XGBoost: objective="binary:logistic", boosting type="gbtree", and the tunned hyperparameters were n estimators [50-1000], learning rate [0.001-0.5], max depth [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20], colsample by tree [0.2-1], and scale pos weight .…”

Section: E Proposed Ml-based Methodologymentioning

confidence: 99%

“…As reviewed in recent survey works [4]- [6], decisionsupport systems based on machine learning (ML) techniques have been proposed for application in emergency medicine. Indeed, in the context of this paper, for EMS, there are many interests in using ML methods for tasks such as: identifying possible medical conditions before arrival on emergency departments [7], to predict ambulances' demand to allow their reallocation [1], to predict ambulance response time [8], [9], to predict the ambulances' turnaround time in hospitals [10], to predict clinical outcomes [11], to early identify clinical conditions on emergency calls [12], to recognize and predict service disruptions [13], and so on.…”

Section: A Backgroundmentioning

confidence: 99%

See 1 more Smart Citation

Privacy-Preserving Prediction of Victim’s Mortality and Their Need for Transportation to Health Facilities

Arcolezi

Cerna

Couchot

et al. 2022

IEEE Trans. Ind. Inf.

Self Cite

View full text Add to dashboard Cite

Emergency medical services (EMS) provide crucial prehospital care, such as in the case of cardiac arrest, where the victim requires immediate first-aid. For this reason, it is vital to improving EMS response time. This paper proposes a novel methodology based on machine learning (ML) techniques to predict both the victims' mortality and their need for transportation to health facilities using data gathered from the start of the emergency call until the Departmental Fire and Rescue Service of the Doubs (SDIS25) is notified. We first analyzed SDIS25 calls to find out associations between the call processing times and victims' mortality, and to measure the variables' importance. Next, we validated our proposed MLbased methodology, where mortality could be predicted with accuracy and area under the receiver operating characteristic curve (AUC) scores of 96.44% and 96.04% respectively, while the need for transportation achieved an accuracy and AUC scores of 73.62% and 78.91%, respectively. What is more, we found out that it was still possible to predict both targets perturbating the input data by applying k-anonymity and differential privacy techniques. In conclusion, the results showed the potential of ML for EMS, which can be used as a decision-support tool to early identify mortality and the use of resources (transportation) and, thus, help EMS to save more lives and avoid service disruptions.

show abstract

“…Because many victims require care within an adequate time frame (e.g., cardiac arrest), improving response time is vital. [3] proposes to predict ARTs using machine learning techniques, which could be used as a decision support system by EMS to enable dynamic selection of ambulance dispatch centers. A well-known predictor of ARTs is the location of the emergency (e.g., whether it is an urban or rural area), which is a sensitive input because it can re-veal who received care and for what reason.…”

Section: Predictions On Anonymized Datamentioning

confidence: 99%

Predictions in Pre-Hospital Emergency Transport in France: A State of the Art

Guyeux¹

2022

Signal Processing and Vision

View full text Add to dashboard Cite

For a number of years now, the regional fire department centers have been recording their interventions numerically. Such databases are under- utilized and are mainly used for statistical andmanagement purposes. However, such a history of interventions can be very useful, if used in conjunction with artificial intelligence algorithms, for predictive purposes. Such work has recently been done in France through a series of articles investigatingthe various aspects of the problem, and has been put into production at the Doubs center. The objective of this review is to take stock of all the workthat has been done so far, to list the successes andthe stumbling blocks, and to draw up a roadmap on this theme for the years to come.

show abstract

Preserving Geo-Indistinguishability of the Emergency Scene to Predict Ambulance Response Time

Cited by 8 publications

References 39 publications

Differentially private multivariate time series forecasting of aggregated human mobility with deep learning: Input or gradient perturbation?

Differentially private multivariate time series forecasting of aggregated human mobility with deep learning: Input or gradient perturbation?

Privacy-Preserving Prediction of Victim’s Mortality and Their Need for Transportation to Health Facilities

Predictions in Pre-Hospital Emergency Transport in France: A State of the Art

Contact Info

Product

Resources

About