Advertisement
Science

Nutrition perfected by machine learning

Advertisement

the National Institute of Nuclear Physics (Infn) Naples section and the Research and Development company Antur srl, have created a completely innovative combination. This is the management and application of data analysis techniques combined with machine learning techniques. The data was collected by ANTUR, machine learning was used to study the characteristics of these samples in order to then be able to divide them into groups (generally called “clusters”) that shared common characteristics with the aim of recreating a nutritional approach for each cluster specifics dedicated. The artificial intelligence used is a map of the Self Organizing Map (SOM) type, which deals with the organization of networked information processes similar to artificial neural networks.

 

Characteristic of the Self Organizing Maps is that they operate using the so-called unsupervised learning ,  i.e. an automatic learning technique that allows the computer system to work on a series of inputs (data) which it will reclassify and organize itself by looking for common characteristics in order to perform reasoning and predictions on subsequent inputs.

Smart nutrition?

The characteristics of the SOM have been exploited to “identify, in the field of nutrition, which foods can also act as a therapeutic tool for the treatment of individual pathologies”, said Maria Luisa Conza, Project Manager and Research Director ANTUR .

“This machine learning, the algorithm thus constructed, by feeding it over time with a much larger quantity than the single data acquired to date through an instrument called Bioimpedancemetry or BIA which evaluates the body composition of the single individual, will be able to make a network intelligent neural to the point of being able to predict, simply by entering biomedical parameters such as weight, height and sex of the patient, the characteristics of an adequate diet capable of solving the pathological problem of the single individual. This system can be used by all operators in the “health” sector, biologists, nutritionists, general practitioners and specialists. My next goal will be to share and implement this project with a group of experts oncologists by activating a transversal study also with the INFN and the CERN studies in Geneva”, these were the declarations of Dr. Conza during the workshop held in Naples in which the project was presented.

 

Advertisement
Advertisement

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button