Deep Learning Like a Viking - Building Convolutional Neural Networks with Keras

Explore the fascinating world of Viking runes and deep learning in this 47-minute Devoxx conference talk. Dive into the Younger Futhark writing system and learn how to build a Convolutional Neural Network using Keras to recognize hand-written runes. Follow along as the speaker demonstrates the creation of a web application with Flask and JavaScript, allowing real-time rune recognition. Gain insights into the workings of Convolutional Neural Networks, their visual hierarchy, and practical implementation using Python and Keras. Discover the potential applications of this technology, from ancient scripts to modern symbols like the Bluetooth logo. By the end of the talk, acquire the knowledge to construct your own neural networks and integrate them into applications using Flask.

Introduction
Im NOT a Data Scientist
Runes
Younger Futhark
The Runes
Bluetooth
Latin alphabets
Egyptian hieroglyphics
Handwritten hieroglyphics
Neural Networks
Bias
Math
Activation
Perceptron
Output
Layers
Recognize Ruins
Convolutional Neural Networks
Visual Hierarchy
Pooling
Flattening
Lego Viking
Theory
Building Neural Networks
What is Keras
Keras Code
Demo
Code
Flask
Build the model
Running the model
Limitations of the model
Resources
Questions