Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Lingua: | |
| Pubblicazione: |
Zenodo
2025
|
| Accesso online: | https://doi.org/10.5281/zenodo.15153707 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
Sommario:
- <p>A huge number of deaths and injuries caused by car accidents is one of the most urgent problems we <br>will face today in the world. Ambulances can be pre-positioned to shorten response time and give immediate <br>medical assistance instead of being dispatched just when needed. When it comes to healthcare decision-making <br>and problem-solving, deep learning approaches have already shown to be indispensable. In order to forecast the <br>best places to position ambulances, this research presents a method based on deep-embedded clustering. It is <br>critical to understand these linkages while creating models because road crash occurrence is heavily influenced <br>by various regional characteristics and patterns. This study uses CAT2VEC, another model of deep learning, to <br>implement the need to maintain such formulas in the entire model building to ensure real -time results. <br>Traditional clustering techniques such as K-Means, GMM and agglomerative clustering are also compared to <br>the proposed framework. For the purpose of evaluating the efficiency of various algorithms, a new feature of <br>scoring to calculate the time and distance of the response in real time has been developed. The proposed <br>ambulance system acts impressively exceeds all other conventional methods, reaches a speed of accuracy of <br>95% with cross validation of K-time and a unique score of 7.581 distance.</p>