Federica Mutti
Curriculum Vitae
PERSONAL DETAILS
Birth February 25, 1992
Address Viale Abruzzi 1, Milan (Italy)
Phone +39-347-*******
Mail ac8go6@r.postjobfree.com
EDUCATION
The Data Incubator 2017
Data Reply IT
An intense eight weeks data science bootcamp, focused on machine learning, big data and data visualization
Master’s degree in statistics for enterprises 2014-2017 University of Milano-Bicocca
Thesis title: "Development of an image classi cation process to identify longitudinal crack on road surface"
Evaluation: 110/110 cum laude
Bachelor’s degree in statistics 2011-2014
University of Milano-Bicocca
Thesis title: "Methods for detecting multivariate outliers: an application to the quality of underground waters"
Evaluation: 110/110 cum laude
WORK EXPERIENCE
Junior Data Scientist 2017-present
Data Reply IT
Crop eld extraction from satellite imagery: a watershed algorithm has been applied on two years satellite tiles from ESA Sentinel 2 in order to extract agricultural crop eld boarders. The algorithm has been based on enriched images instead on raw ones: the Normalized Di erence Vegetation Index (NDVI) has been calculated on each pixel location of each image.
Costumer: farm management solutions provider in agriculture industry Location: Milan
Duration: January 2019 - present
Tools:: Python for modeling
Control room implementation: simple-threshold-based triggers, univariate and multi- variate anomaly detection algorithms based on fuel consumption, speed and temperature data have been developed in order to monitor the usage of some agriculture vehicles during their harvesting period.
Costumer: automotive company
Location: Turin, Milan, Chicago
Duration: November 2017 - present
Tools:: Hive and Impala for ETL, R and Python for modeling, Tableau for visualization
Optimizing pro ts from hydroelectricity production: a mathematical model design- ing the hydroelectric complex of interest has been de ned in order to nd a 24-hours production plan, for each power plant along the hydroelectric complex, that maximizes the pro ts resulting from electricity.
Costumer: energy company
Location: Milan
Duration: July 2018 - October 2018
Tools:: Python for modeling
Planned maintenance alerting: ne-tuning of a model performing an engine hours prediction for some commercial vehicles in the next 6 weeks and so determining their future maintenance steps. An ARIMAX or an Exponential Smoothing State Space model has been used to compute the daily engine hours increments, according to the historical data of each vehicle.
Costumer: automotive company
Location: Milan, Turin
Duration: July 2017 - October 2017
Tools:: Hive and Impala for ETL, R for modeling
CCT prediction: a neural network autoregression with a single hidden layer has been implemented to predict the value that CCT (an Italian energy price index de ning as the di erence between the zone price of energy and its national price) will assume in the next month in the North part of Italy.
Costumer: energy company
Location: Milan
Duration: May 2017 - June 2017
Tools:: R for modeling
SKILLS
Languages Italian (mother tongue), English (
uent) Application
Platforms
Hadoop (Impala, Hive), Microsoft Azure, DataBricks Programming
Languages
Python, R, SQL, SAS
Operating
Systems
Windows, macOS
Other RStudio, Jupyter Notebook, Spot re, Qlik Sense, Tableau