University Projects
Dec 2020 - Feb 2021
Designed and implemented several Deep Learning models for classification of malignant and benign masses and calcifications achieving up to 87% accuracy.
WHAT DID I LEARN?
• Put into practice various techniques for Deep Learning
• Trained my ability to work with heterogeneous teams and in English language
Brain Based Wheel Chair Control
We have realized the process of making a wheelchair controlled through the use of a helmet and an onboard computer capable, through Machine Learning systems, of translating the user's thoughts into movements to be made by the chair. The process was defined in terms of Machine Learning Pipeline, Use Case analysis, cost assessment and technical feasibility. We have created, using BPMN language, the service diagrams of the various steps and have also simulated various cases through the use of BIMP. During the process, we applied process mining, heuristic, random and inductive techniques using ProM and Apromore and we made evaluations on the mined models and on the logs extracted from BIMP.
WHAT DID I LEARN?
• To study and build a Machine Learning Pipeline
• How to perform Handoff and service-level diagrams using BPMN language
• Simulate and evaluate processes using BIMP
• Apply process mining techniques using Apromore and ProM
Implementation of the KMeans clustering algorithm through the use of the MapReduce distributed programming framework. The result was two applications, one developed with Java and Hadoop and the other with Python and Spark. The programs were aimed at clustering different types and quays of data and their performance was evaluated through statistical analysis carried out on datasets heterogeneous in size, size and number of clusters.
WHAT DID I LEARN?
• Correct way to develop distributed applications in a cloud environment
• Use of Hadoop, Spark and in general of the Map Reduce framework.
Smart Home IoT
Creation and simulation of an IoT application for the realization of a smart home using Cooja, Contiki and Californium (Java). The project focused on monitoring energy consumption and temperature in the various rooms of the house. This information was then used by the system to control the windows and optimize the energy efficiency of the house.
WHAT DID I LEARN?
• Learned how to use Contiki, Cooja and Californium
• Experienced for the first time, the implementation of an IoT environment
• Java application using MongoDB database for the monitoring of past natural disasters
• Java application using Neo4j graph database to track publications scraped from Google Scholar using Scholarly and Python
• Java application using Java Persistence API (JPA) and key-value database for the management of a bookshop
WHAT DID I LEARN?
• Improved my Python and Java skills
• Learned to use NoSQL databases (Document Databases with MongoDB, Key-Value databases with LevelDB and Graph Databases with Neo4j)
• Experimented with JPA framework and learned the approaches to grant persistence in an application
• Learned basic concepts and techniques of web scraping with Python.
Desktop application for the managing of an intelligent fridge. I worked on the project from analysis to design and development using Java, SQL, UML and XML.
WHAT DID I LEARN?
• Learned the Java fundamentals and to use UML and XML
• Learned to use agile methodologies to develop software in a better way
Relational Database for Restaurant Chain
A relational database designed for a very wide restaurant chain. I used SQL applying knowledge for the normalization of the database and Dia to design the E-R model.
WHAT DID I LEARN?
• The knowledge needed to build a very large and complex relational database
Browser game set in the Vikings' epoch developed using HTML, CSS, Javascript and PHP.
WHAT DID I LEARN?
• Improved my Web Development skills
• Tested me to think of a completely new complex game from scratch maintaining it functioning and consistent
Last updated