/GlobalAIHubDLCourse

Repository for exercises and project of the intro to deep learning course in Global AI Hub

Primary LanguageJupyter NotebookMIT LicenseMIT

GlobalAIHubDLCourse

LO

Course Date: March 1 to March 5, 2021

Name: Nuria Andrea

Surname: Malet Quintar

Email: nmalet12@gmail.com

Project:

Image classification project using CIFAR-100 for training. First I create a CNN with Keras, fit the model and use it to predict labels for 5 images downloaded from internet. Then downloaded the vgg16 model and use it to predict the same 5 images. Finally, I retrained the classifier part of the vgg16 with the CIFAR-100 dataset, keeping all other layers as in the original model, then use it to classify the same 5 images. The best results were achived in the last case.

Requirements

Python Libraries

  • numpy
  • pandas
  • tensorflow/keras
  • (for the first homework there are a lot of libraries needed to read different kinds of data)

Certification

Top Learner Certificate