Basics of Neural Networks

Cambodia Crop Mapping

This page contains all slides and agenda for the training.

Agenda

Day	Content
Day 1: 05/25, 9 - 11 am	Where are we? Summary of the crop mapping work Introduction to Artificial Intelligence (AI) and Machine Learning (ML) Types of AI and ML Introduction to deep learning & Neural Networks
Day 2: 05/26, 8:30 - 10:30 am	Recap Day 1 Most common model structures (Neural Networks) Terminology Advantages and Disadvantages
Day 3: 05/27, 8:30 - 10:30 am	Recap Day 2 “Play” with Neural Network demos Main discussion about the crop mapping work
Day 4: 05/28, 8:30 - 10:00 am	Recap of previous days Working in the environment: Brief Introduction to Google Machine Learning, Google Colab, Google Cloud Platform and TensorFlow Next steps

Materials

• Day 1: Slides 1 | Recording 1

• Day 2: Slides 2 | Recording 2

• Day 3: Slides 3 | Recording 3

• Day 4: Slides 4 | Recording 4

Resources

• How Machine Learning Works by Yann LeCun

• Medium article “Machine Learning in 8 minutes”

• Medium article Towards AI “What is Machine Learning”

• Machine Learning Mastery article “What is Semi-Supervised Learning”

• Google Quick Draw game

• YouTube video “Dance Moves of Deep Learning Activation Functions”

• Towardsdatascience article “A Comprehensive Guide to Convolutional Neural Networks”

• TensorFlow Playground

• TensorFlow page

• Google Colaboratory FAQs

• Development Seed “Sattelite Machine Learning Training”

• Activation Functions in Neural Networks, Towardsdatascience article

• Google’s Machine Learning Crash Course

• TensorFlow YouTube video Zero to Hero

• Michael Nielsen’s Neural Network and Deep Learning content

• Google Earth Engine tutorials

• End-to-end Google Earth Engine by Ujaval Ghandi

Acknowledgement

This was developed by SIG, SIG-NAL through funding with SilvaCarbon for use in the Cambodia crop mapping work.

These materials build on work developed and supported by SERVIR and Google TensorFlow workshops.