Advanced statistics: AI & Machine Learning - Prof. Marcello Mascini - a.a. 2023/2024
Food science and technology - Elective
![notification_alert_warning.png](https://elearning.unite.it/local/notification_alert_warning.png)
Topic outline
-
Marcello Mascini is associate professor in analytical chemistry. In his academic career Prof Mascini has published more than 100 papers (70% as corresponding author) on international scientific journals (peer-reviewed) with an average Impact Factor of 4.2.
He actively participated in national and international scientific meetings with more than 200 posters or oral presentations (abstracts). Prof. Mascini research area was focused on the development of screening methods for fast and real time detection of analytes important for health and quality control analysis. The research interests were with a particular focus on new methods to develop bio-synthetic systems (biomimetic or bioinspired) in analytical application using molecular modeling and advanced multivariate system
The course is split in 2 units
Unit 1: Clustering and Classification (2 CFU).
Nested k-fold cross-validation PLS; K-means clustering; Gaussian mixture model, classification and regression trees; K-nearest neighbor; naive bayes classifier; Association rules, Non-Linear models
Unit 2: support-vector networks and Neural networks (2 CFU).
Support-vector machine and multi layer neural network. network validation. Network test and query with external data.The course provides an in-depth study of multivariate statistics starting from the knowledge acquired during the first year of the this degree. In particular, the concepts related to clustering, classification and neural networks will be studied using data from literature.
The main objective is to provide an in-depth application of statistics by providing tools to correctly choose the tests to process "big data".
The literature data will be processed with free software of recognized international quality (R software, justNN).
The skills acquired for statistical treatment of experimental data will assist in professional career.
Because of the practical application nature of this course there is no mandatory textbook. Instead, you should purchase a text that suits your needs (e.g., practical application versus mathematical statistics). Recommended texts are:
slides of the lessons and eLearning tools.
Slides of the lessons and eLearning online tools
Date: at the end of the unit (please see the planning)
EVALUATION
One oral presentation in form of report
The Teacher is available to answer questions at the end of the lesson, or on request by e-mail (mmascini@unite.it)
-
-
The Link to partecipate to the students reception Online. Please write me an email to book an appointment:)
-
-
-