This repo shows some small projects done for the Machine Learning course during my Informatics and Computer Engineering Masters specializing in Intelligent Systems and Data Analysis and Processing at IST.
In Homework 3 our group implemented by hand basic neural networks using different activation functions and loss functions and then compared the results with the implementations from packages such as sci-kit.
in Homework 5 our group implemented a model capable of recognizing fruits in images. To do so we used transfer learning, where we chose a pre-trained CNN architecture from Keras, froze the earlier layers and trained the last layers with our dataset.