In an automotive context, reliability models are used to study and analyze the probability of failure of the components of the vehicles. However the actual analyses are limited to standard statistical models taking into account only endogenous factors.
The aim of the work is to leverage multiple data sources and combine survival models with more evolved machine learning techniques (e.g. XGBoost with Accelerated Failure Time models).
The results of the models will be integrated into the corporate systems and allow to take better decision in term of aftersales pricing and provisions forecasting.
Data Preparation, Feature Engineering, Statistics, Machine Learning, Python, Cloud Environments
KPMG offre l'opportunità di uno stage curricolare di 4 mesi con supporto per la compilazione della tesi di laurea.
Inizio progetto previsto per aprile 2022.
VI CHIEDIAMO DI ALLEGARE IL CV ALLA VOSTRA CANDIDATURA. Le candidature senza Cv non verranno prese in considerazione.