29 Nov 2017
Completed 12 Dec 2017 (investigating only)
University of Sydney will open a new course: COMP5329 Deep Learning in semester one, 2018. Without a doubt, Deep Learning is going to dominate the trend in 2018, building on from the 2017 trend, machine learning.
TechLab starts to investigate some open-sourced deep learning projects and this is one of them: applying Evolutionary Artificial Neural Networks with Unity. So far, the investigation includes training the Neural Network by switching the activation functions (sigmoid/tanh/softsign functions).
TechLab did some enhancements on this GitHub open-source project and tried to investigate:
Neural network modelling; learn how the Evolutionary Algorithm works.
Is it possible to develop and train a neural network project using framework and languages other than TensorFlow and Python?
Is it possible to use Unity for the data visualisation of machine learning / deep learning training?
The investigation in TechLab had met its goal. The original project had also been enhanced (the activation functions) based on the knowledge from Stanford CS231n course (http://cs231n.github.io/) and deeplearning.ai courses (https://www.coursera.org/learn/neural-networks-deep-learning).
.NET framework, C#, Unity
Github Enterprise Repo
 The SoftSign function as proposed by Xavier Glorot and Yoshua Bengio (2010): “Understanding the difficulty of training deep feedforward neural networks” http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.207.2059&rep=rep1&type=pdf
 CS231n - Convolutional Neural Networks for Visual Recognition: Architecture http://cs231n.github.io/neural-networks-1