The main goals are: review the state-of-the-art of data mining methods for predictive maintenance problems; perform a data exploratory analysis of the sensors transmitted data; apply different learning strategies for anomaly detection; study suitable metrics to assess the effectiveness of models in the early failure detection; diagnose the faults that caused the failure and study the possibility of optimizing the process of detecting the anomalies.
PhD in Computer Science or related areas.
Minimum profile required
Strong background on data mining with good scientific and programming skills.
Knowledge of techniques related to the detection of anomalies and learning from imbalanced data domains. Knowledge of R and/or Python.
Since 07 Sep 2020 to 25 Sep 2020
Cluster / Centre
Computer Science / Artificial Intelligence and Decision Support