Deep learning and neural networks are among the most popular buzz words of the moment. One of the more popular applications of neural networks is Google's Deep Dream, which trains a convolutional neural network on a large image data set and produces the most stunning (and disturbing) concepts. In this project, we aim to apply the methods used in Deep Dream to our own code base in order to discriminate between beautiful and ugly code. We will investigate the workings of the mathematical abstractions behind neural networks and get our hands dirty on building and training our own convolutional neural net.
Approach
The goal is to understand the workings and challenges of training a deep neural network and to build a neural net that is able to classify clean and dirty code using the most famous machine learning toolkit out there: TensorFlow. An important aspect is to get better insight into the possibilities of machine learning and artificial intelligence, and we wish to apply it to some of the unique data available to us.
Joining
We welcome everyone who is interested in learning more about and neural networks, who is not daunted by mathematics, and who doesn't mind spending some time learning the basics of machine learning, Python and Tensorflow before the start of the exploration day. Our goals are quite challenging, so preparation will be vital if we want to get some results.
Languages
Python
Useful material
Artificial Intelligence Courses on Youtube: https://www.youtube.com/user/aicourses
Stanford deep learning tutorial: http://deeplearning.stanford.edu/tutorial/
Tensorflow's convolutional neural network tutorial: https://www.tensorflow.org/versions/master/tutorials/deep_cnn
Built With
- artificial-intelligence
- machine-learning
- neural-networks
- python
Log in or sign up for Devpost to join the conversation.