ELEKTRİK-ELEKTRONİK MÜHENDİSLİĞİ BÖLÜMÜ


Derin Öğrenme: Uygulamalı Eğitim

 
A joint activity of 
METU Electrical and Electronics Engineering
METU-HU Neuroscience and Neurotechnology Ph. D. Program
Intel Nervana AI Academy METU Branch
METU VISION Research Lab
 
Tutorials by:
Egemen SERT, Barkın TUNCER, Sefa Burak OKCU, Eren HALICI, Fourough GHARBALCHI
 
Mentor:
Ugur HALICI, Prof. Dr.
 
A 12-week-long Deep Learning tutorial series including workshops for practice will be initiated on 
October 25, 2017 Wednesday (18:40-20:30 at U3). 
The tutorial on October 25 will be held at 19.00 due to maintenance on U3.

The topics starts from the Deep Learning Basics and will cover a broad range. 
 
For participants, familiarity with Multivariate Calculus, Linear Algebra and Python would be a plus; 
however, related topics will be covered in the class, hence familiarity is not a prerequisite.

Throughout the turotial series, participants will learn how to build object classifiers, text-to-text translators, 
Style Transfer models (https://github.com/fzliu/style-transfer) and so on.

Participants are required to fill the participation form. Students are welcome.
https://goo.gl/forms/kADnzzZvLqBlof6S2

If you filled the form, please join our Deep Learning Community via the link below.
https://deepmetu.slack.com/signup

For workshops, bringing computers is a must; however, for regular classes, computers are not required.

The tentative topics for each week can be found below.

1. Introduction: What is Deep Learning? Why do we need data? 
            What can be done with DL? What is loss function intro
 
2. Neural Networks: What is regression, classification? 
            What is Neural Network, What does the Neural Network learn?
 
3. Neural Networks Training: What is Backpropagation? What is Gradient Descent? 
            How does the Neural Network learn?
 
4. WORKSHOP I: NN implementation, SGD, Dropout, Batch Normalization, Adam Optimizer,
            Train/Test Splits, Training CIFAR-10
 
5. Convolutional Neural Networks: What is Convolution? What is kernel/filter? What is pooling? 
            What is dilation? What is segmentation?
 
6. CNN Architectures: AlexNet, VGGNet, ResNet, GoogLeNet, Highway Net, UNet, DenseNet, 
            R-CNN, Faster-R-CNN…
 
7. WORKSHOP II: segmentation example? face detection example? (with Keras Framework)
 
8. Recurrent Neural Networks: What is an RNN? Why do the gradients vanish? How stable is vanilla RNN? 
            What is gating? What is LSTM & GRU?
 
9. Attention: What is Attention? What is External Memory? Neural Machine Translation, Image Captioning (w/ attn)…
 
10. WORKSHOP III: English-French translation, video search, image captioning (with Keras Framework)
 
11. Autoencoders: What is AE, GAN, VAE, Style Transfer
 
12. Deep Reinforcement Learning