The following work aims to examine and compare some statistical techniques and models in the field of image classification. Particular emphasis is given to the deep approach that characterizes deep neural networks and the advantages of using this type of model with respect to the problem under consideration. Each model is defined by starting from a theoretical basis and then applying it to a real dataset and finally comparing the results obtained with other models. Specifically, a brief introduction to deep learning and neural networks is given in Chapter 1. In Chapter 2, on the other hand, we describe the models under consideration, which are then, in Chapter 3, applied to a dataset of biological cell images and compared with respect to the higher accuracy obtained in classifying each cell. The analyses are carried out using the R and Python languages.
Il lavoro seguente ha l’obiettivo di esaminare e confrontare alcune tecniche e modelli statistici nell’ambito della classificazione di immagini. Particolare rilevanza viene data all’approccio deep che caratterizza le reti neurali profonde e ai vantaggi che derivano dall’utilizzo di questo tipo di modello rispetto al problema in esame. Ciascun modello viene definito partendo da una base teorica per poi applicarlo ad un dataset reale ed infine confrontare i risultati ottenuti con gli altri modelli. In particolare, nel capitolo 1 si fornisce una breve introduzione al deep learning e alle reti neurali. Nel capitolo 2, invece, si descrivono i modelli in esame che poi, nel capitolo 3, vengono applicati ad un dataset di immagini di cellule biologiche e confrontati rispetto alla maggior precisione ottenuta nella classificazione di ogni cellula. Le analisi sono svolte usando i linguaggi R e Python.
Deep learning per la classificazione di immagini: un approfondimento sulle reti neurali profonde
LOLLATO, FRANCESCO
2022/2023
Abstract
The following work aims to examine and compare some statistical techniques and models in the field of image classification. Particular emphasis is given to the deep approach that characterizes deep neural networks and the advantages of using this type of model with respect to the problem under consideration. Each model is defined by starting from a theoretical basis and then applying it to a real dataset and finally comparing the results obtained with other models. Specifically, a brief introduction to deep learning and neural networks is given in Chapter 1. In Chapter 2, on the other hand, we describe the models under consideration, which are then, in Chapter 3, applied to a dataset of biological cell images and compared with respect to the higher accuracy obtained in classifying each cell. The analyses are carried out using the R and Python languages.File | Dimensione | Formato | |
---|---|---|---|
Lollato_Francesco.pdf
accesso aperto
Dimensione
1.02 MB
Formato
Adobe PDF
|
1.02 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/52445