A diary about my journey to explore Deep Learning

Getting Started

In June 24, I started my journey with Deep Learning. At first day, I started with deeplearning.net, but it didn’t work very well. Another resource, Deep learning tutorial makes me more confuse when I saw some complex mathematical notations. I found a good slide I feel comfortable with A tutorial on deep learning at icml 2013 of Yann Lecun. At second day, I found Andrew’s course about Deep Learning – Unsupervised Feature Learning and Deep Learning at openclassroom. I want to solve simple logistic regression and simple neuron network by myself. At 4th day, I found Deep Learning – Methods and Applications book of Li Deng and Dong Yu from this quora question

Interesting courses (5th day)

At 5th day, I found 2 interesting courses Document-icon CS224d: Deep Learning for Natural Language Processing and camera CS231n Convolutional Neural Networks for Visual Recognition From course CS224d, I found the first breakthrough result of deep learning in speech recognitionContext-Dependent Pre-Trained Deep Neural Networks for Large-Vocabulary Speech Recognition by Dahl et al. (2010), breakthrough in computer vision  Imagenet classification with deep convolutional neural networks by Krizhevsky et al. (2012)

GTX 980 (21th day)

Deep learning requires a lot of matrix computation, it leads to requirement that we must have a GPU, according to Andrew Ng in his talk at GPU Technology Conference 2015

With this help of this article, I choose GTX 980 for my research.

First Challenge: Facial Keypoints Detection (22th day)


I follow this awesome tutorial of Daniel Nouri to use my GPU and do this challenge


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s