Venice is drowning

a Streaming Data Management and Time Series Analysis project

Overview  | Data sources  | Objectives  |   About us  

☍   Overview

An attempt to model the tidal phenomenon in the Venice lagoon using both linear/statistical and machine learning approaches.

☍   Data sources

• kaggle dataset
• github repo inspiration
• tides data for 2018
• ARPA Veneto for meteorological data
• PyEphem API for lunar motion
• oce package for Analysis of Oceanographic Data

☍   Objectives

The main objective of the project is to analyze the data of the tide detections regarding the area of the Venice lagoon, producing predictive models whose performances are evaluated on a time horizon ranging from one hour up to a week of forecast.

For this purpose, three models, both linear and machine-learning based, are tested:

• ARIMA (AutoRegressive Integrated Moving Average);
• UCM (Unobserved Component Models);
• LSTM (Long Short-Term Memory).

The final report is available here while here the slides.

☍   About us

⊜   Dario Bertazioli

• Current Studies: Data Science Master Student at Università degli Studi di Milano-Bicocca;
• Past Studies: Bachelor's degree in Physics at Università degli Studi di Milano.

⊜   Fabrizio D'Intinosante

• Current Studies: Data Science Master Student at Università degli Studi di Milano-Bicocca;
• Past Studies: Bachelor's Degree in Economics e Statistics at Università degli Studi di Torino.