Skip to content

Latest commit

 

History

History
13 lines (10 loc) · 752 Bytes

README.md

File metadata and controls

13 lines (10 loc) · 752 Bytes

instagramlikeprediction

Instagram Like Prediction with Deep Learning using Keras

Project work on Intelligent Systems M at University of Bologna

The objective of this project is predicting the number of likes on Instagram using Deep Learning. I've made various attempts trying to perform predictions better and better. I started using meta-data only (n° followers, n° posts, hashtags, ...), then using images (with Convolutional Neural Networks) and finally combining both images and labels acquired using Google Vision API.

All the project is written in Python 3. I've used Keras as deep learning framework. The folder 'Tutorials' contains my first experiments with Python and Keras. See the presentation slides for a more detailed explanation.