Search results
1 – 4 of 4Ali Sadeghi-Naini and Ali Asgary
A feed-forward back-propagation neural network (NN) is proposed to model number of firefighters responding to different fire incidents. Such a predictor model can estimate number…
Abstract
Purpose
A feed-forward back-propagation neural network (NN) is proposed to model number of firefighters responding to different fire incidents. Such a predictor model can estimate number of firefighter personnel required to tackle new incidents. This a priori information at the time of dispatch can help saving unnecessary efforts in low-risk incidents while focussing on high-risk ones to reduce overall damages and injuries caused by the fire incidents.
Design/methodology/approach
A fully connected multilayer NN was adapted as the prediction model. The network was trained on a large number of fire incident records reported in Toronto area between 2000 and 2006 and then its performance was evaluated on another set of never seen records. Two types of prediction were done to model number of responding personnel: a rough category prediction and an exact number prediction.
Findings
Results obtained reported a very promising ability of this approach to model number of firefighters responding to a fire incident.
Originality/value
Such a model can significantly reduce uncertainties on the requirements needed for tackling a fire incident once it is reported.
Details
Keywords
Hasnae Zerouaoui, Ali Idri and Omar El Alaoui
Hundreds of thousands of deaths each year in the world are caused by breast cancer (BC). An early-stage diagnosis of this disease can positively reduce the morbidity and mortality…
Abstract
Purpose
Hundreds of thousands of deaths each year in the world are caused by breast cancer (BC). An early-stage diagnosis of this disease can positively reduce the morbidity and mortality rate by helping to select the most appropriate treatment options, especially by using histological BC images for the diagnosis.
Design/methodology/approach
The present study proposes and evaluates a novel approach which consists of 24 deep hybrid heterogenous ensembles that combine the strength of seven deep learning techniques (DenseNet 201, Inception V3, VGG16, VGG19, Inception-ResNet-V3, MobileNet V2 and ResNet 50) for feature extraction and four well-known classifiers (multi-layer perceptron, support vector machines, K-nearest neighbors and decision tree) by means of hard and weighted voting combination methods for histological classification of BC medical image. Furthermore, the best deep hybrid heterogenous ensembles were compared to the deep stacked ensembles to determine the best strategy to design the deep ensemble methods. The empirical evaluations used four classification performance criteria (accuracy, sensitivity, precision and F1-score), fivefold cross-validation, Scott–Knott (SK) statistical test and Borda count voting method. All empirical evaluations were assessed using four performance measures, including accuracy, precision, recall and F1-score, and were over the histological BreakHis public dataset with four magnification factors (40×, 100×, 200× and 400×). SK statistical test and Borda count were also used to cluster the designed techniques and rank the techniques belonging to the best SK cluster, respectively.
Findings
Results showed that the deep hybrid heterogenous ensembles outperformed both their singles and the deep stacked ensembles and reached the accuracy values of 96.3, 95.6, 96.3 and 94 per cent across the four magnification factors 40×, 100×, 200× and 400×, respectively.
Originality/value
The proposed deep hybrid heterogenous ensembles can be applied for the BC diagnosis to assist pathologists in reducing the missed diagnoses and proposing adequate treatments for the patients.
Details
Keywords
Maryam Shafiei Sabet, Ali Asgary and Adriano O. Solis
Responding to emergency incidents by emergency response organizations such as fire, ambulance and police during large disaster and emergency events is very important. The purpose…
Abstract
Purpose
Responding to emergency incidents by emergency response organizations such as fire, ambulance and police during large disaster and emergency events is very important. The purpose of this paper is to provide some insights into response patterns during the 2013 ice storm in the city of Vaughan, Ontario, Canada, using temporal and spatial analyses.
Design/methodology/approach
The City of Vaughan Fire and Rescue Service data set containing all responses to fire and other emergency incidents from January 1, 2009 to December 31, 2016 was used. The 2013 Southern Ontario ice storm occurred from December 20, 2013 to January 1, 2014, and, for this study, December 20–31 is considered the “study period.” Temporal, spatial and spatiotemporal analyses of responses during the study period are carried out and are compared with the same period in other years (2009–2012 and 2014–2016).
Findings
The findings show that temporal patterns of response attributes changed significantly during the 2013 ice storm. Similarly, the spatial pattern of responses during the 2013 ice storm showed some major differences with other years. The spatiotemporal analyses also demonstrate significant variations in responses in the city during different hours of the day in the ice storm days.
Originality/value
This study is the first study to examine the spatiotemporal patterns of responses made by a fire department during the 2013 ice storm in Canada. It provides some insights into the differences between response volumes, temporal and spatial distributions during large emergency events (e.g. ice storm) and normal situations. The results will help in mitigating the number of responses in the future through public education and technological changes. Moreover, the results will provide fire departments with information that could help them prepare for such events by possible reallocation of resources.
Details