| Date | Syllabus | Homework |
|---|---|---|
| 3/1 | Course Overview | |
| 3/7 | HW1-1: Introduction to Python & Data | |
| 3/8 | Introduction to NLP & Deep Learning | |
| 3/14 | HW1-2: Introduction to TensorFlow & Data | |
| 3/15 | Backpropagation | |
| 3/21 | HW1-3: Backpropagation | |
| 3/22 | Classic Representations & Language Modeling | |
| 3/28 | HW2: Introduction to Keras & Data | |
| 3/29 | Neural Language Modeling & Word Embeddings | |
| 4/4 | Break | |
| 4/5 | Break | |
| 4/11 | HW3: Word Embeddings & Recurrent Neural Networks & Data | |
| 4/12 | Topic Models & Recurrent Neural Networks | |
| 4/18 | Break | |
| 4/19 | Break | |
| 4/25 | HW4: Convolutional Neural Networks & Data | |
| 4/26 | Convolutional Neural Networks | |
| 5/2 | Discussion on Homework 4 & HW5 Preparation | |
| 5/3 | Paragraph Embeddings & Attention | |
| 5/9 | Break & Submit Your Member List! | |
| 5/10 | Advanced Structures | |
| 5/16 | HW5: Generative Adversarial Network & Data | |
| 5/17 | Cycle GAN & Conditional GAN | Submit Your Research Topic |
| 5/23 | IRGAN | |
| 5/24 | Speech Recognition with Deep Learning Methods (Mr. Ming-Han Yang, Delta Electronics, Inc.) | |
| 5/30 | Discussion on Homework 5 | |
| 5/31 | Machine Comprehension with Deep Learning (Mr. Wen-Chin Huang, NTU & Academia Sinica) | |
| 6/6 | Break for Your Final Project | |
| 6/7 | Break for Your Final Project | |
| 6/13 | Break for Your Final Project | |
| 6/14 |
7. ³¯o®õ ªLo§D ¤ýoµ¾¡G Dank Learning: Generating Memes Using Deep Neural Networks 29. ³¯o§» ªòoÀ¯¡G An Extended Sentiment Dictionary Approach with TF-IDF to Enhance Sentiment Analysis Taking the Dcard Job forum as an example 25. ¥]o§t »¯o¥à¡G Very Deep Convolution Networks for Large-Scale Image Recognition 30. ´¿oºÓ¡G Deep Music Genre 10. ¤Bo¦w §do°a §õo®¦¡@ 17. ±io»¨ ´öo¶W ªLoõ¡G Dimensional Sentiment Analysis Using a Regional CNN-LSTM Model 11. ±ço»í ¿½o¤è¡G Explaining and Harnessing Adversarial Examples¡@ 5. §õo²» ±io¦ö ¾Ho¦t¡G Deep Learning for Event-Driven Stock Prediction |
|
| 6/20 |
1. §fo¿Î ©Po»¨ ®}o¨°¡G Recognition of Pornographic Web Pages by Classifying Texts and Images 3. ¼ïo¸Û ½²oÀM ªôoÛ¡G Visual Attribute Transfer through Deep Image Analogy 2. ®ÉoËÕ §do¾ì Ùo¤t¡G SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size 8. ¾GoÁ¾ ³\oÁ¾ ªôo¿Ä¡G LSTM-based deep learning models for non-factoid answer selection |
|
| 6/21 |
20. ¸o»¨ ¦¿oºû¡G MobileNetV2: Inverted Residuals and Linear Bottlenecks 27. ªôoÛ ½²oÀM ¼ïo¸Û¡G Visual Attribute Transfer Through Deep Image Analogy 9. ±io¯E ³¯oÊo ¬Io¥õ¡G Using Neural Networks to Predict Emoji Usage from Twitter Data 19. ¸o«Å §do®ß¡G Poker-CNN a pattern learning strategy for making draws and bets in poker games using convolutional networks 13. ªôo´¼ ³¯oºÍ¡G Playing Atari with Deep Reinforcement Learning 4. ªLo´@ ±io¶® §foµÓ¡G Sequence to Sequence Weather Forecasting with Long Short-Term Memory Recurrent Neural Networks 23. ¨Ho¤@ ½²o¼w¡G Car Plate Recognition Based on CNN Using Embedded System with GPU 28. ®}o¶v¡G Stacked Hourglass Networks for Human Pose Estimation |
|
| 6/27 |
18. ±ioªF ¸âo§Ê¡G Building Energy Consumption Prediction: An Extreme Deep Learning Approach 6. ³¢o©ú ¶Ào´@ §ºo¸R¡G A Deep Learning Approach for Intrusion Detection Using Recurrent Neural Networks 22. ¶Ào¤¯ ¿co¨ç §do§»¡G Generating Chinese Classical Poems with RNN Encoder-Decoder 12. ³¯oµ¾ ¼Bo±j¡G Comment Abuse Classification with Deep Learning¡@ |
|
| 6/28 |
15. §Eo¯E §foèû §doºÂ¡G Sentiment Analysis on Movie Reviews using Recursive and Recurrent Neural Network Architectures 24. ªLo¿A ³¢oèx ·¨o²a¡G Extractive Summarization using Deep Learning 21. ³¯o¿o ±io³Ç¡G Single-Image Crowd Counting via Multi-Column Convolutional Neural Network 14. ¶ÀoËR §õo¦¨¡G Very Deep Convolutional Networks for Text Classification 16. ´¿oµ¾ ¤ýo´P¡G Learning a Wavelet-like Auto-Encoder to Accelerate Deep Neural Networks 26. §õoºü §doÀM ¹ùoµ{¡G Mask R-CNN for Object Detection and Segmentation |