Masterclass follow-up

This is the follow-up for today's (08.02.2020) masterclass on Multi-Class Image Classification.

This repository is divided into 2 parts:

1. Source Code
2. The presentation Python AI Masterclass

Source Code

The given project source codes are written in the easiest way so that most of the audience can digest it and it is strongly advised not to use these configurations on production servers for any type of prediction problems.

• requirements.txt use this file to install all the required packages to proceed further
• config.py is for storing all the configuration parameters and options needed for training and running of the model
• model.py is for generating CNN ClassificationNet model
• train.py is for training ready model
• run.py is for running trained models
• v_data where all the dataset is stored
• classification_model_accs.png the last history plot visualization for accuracy checking from all of the epochs
• classification_model_losses.png the last history plot visualization for loss checking from all of the epochs

Presentation

Presentation is more theoretical to explain the gap between Artificial Intelligence and Multi-Class Image Classification Problem. Also there are definitions for some crucial components that play important role in building custom Neural Networks.