• 沒有找到結果。

National Taiwan University, 2017 Fall Semester Instructor Information

N/A
N/A
Protected

Academic year: 2022

Share "National Taiwan University, 2017 Fall Semester Instructor Information"

Copied!
3
0
0

加載中.... (立即查看全文)

全文

(1)

Page 1

Applied Deep Learning

Machine Learning and Having It Deep and Structured

National Taiwan University, 2017 Fall Semester

Instructor Information

Instructor Email Lecture Location & Hours

Yun-Nung (Vivian) Chen 陳縕儂 Hung-Yi Lee 李宏毅

[email protected]

[email protected] Monday 14:20-17:20 Thursday 09:10-12:10

General Information

Description

Learning the basic theory of deep learning and how to apply to various applications

Expectations and Goals

The students are expected to understand 1) how deep learning works

2) how to frame tasks into deep learning problems

3) how to use deep learning toolkits to implement the designed models, and 4) when and why specific deep learning techniques work for specific problems.

Assignments & Project

4 assignments, 1 group final project, 1 write-up report

Pre-requisites

Required Coding Skills

 Proficiency in Python; All assignments will in Python.

Required Courses

 College Calculus

 Linear Algebra

 Probability

 Machine Learning (preferred but optional)

 Statistics (preferred but optional)

(2)

Page 2

Course Schedule

Week Date Topic Assignment

1 09/11/2017

09/14/2017 Introduction 09/16/2017 Basics

2 09/18/2017

09/21/2017 Neural Networks

3 09/25/2017

09/28/2017 Backpropagtion + Sequence Modeling

4 10/02/2017

10/05/2017 Recurrent Neural Network 10/07/2017 Sentence Completion Release 10/09/2017 Break

5-1 10/12/2017 Company Workshop (Microsoft)

5-2 10/16/2017 Word Embeddings + Guest Lecture (HTC) 6-1 10/19/2017 Guest Lecture (HTC)

6-2 10/23/2017 Word Embeddings + Guest Lecture (Su)

7 10/26/2017 Word Embeddings 10/28/2017 Sentence Completion Due

10/28/2017 Image Captioning Release

8 10/30/2017

11/02/2017 Gating Mechanism + Attention & Memory

9 11/06/2017

11/19/2017 Convolutional Neural Networks 10 11/13/2017 Company Workshop (Microsoft)

11 11/16/2017

11/20/2017 NN Practical Tips 11/19/2017 Image Captioning Due 11/19/2017 Game Playing Release

12 11/23/2017

11/27/2017 Deep Reinforcement Learning 13 11/30/2017 Guest Lecture (Dr. Gao)

14 12/04/2017

12/07/2017 Deep Reinforcement Learning

15 12/11/2017

12/14/2017 Unsupervised Learning 12/16/2017 Game Playing Due

12/16/2017 Comics Generation Release 12/18/2017 Break

16 12/21/2017

12/25/2017 Generative Adversarial Network

17 12/28/2017

01/08/2018 Generative Adversarial Network 01/13/2018 Comics Generation 01/01/2018

01/04/2018 Break

(3)

Page 3

Week Date Topic Assignment

18 01/11/2018

01/15/2018 Final Project Presentation 01/20/2018 Final Project Due

Assignment Schedule

Release Date Due Date Subject Time

10/07/2017 10/28/2016 Sentence Completion 3 weeks

10/28/2016 11/19/2016 Image Captioning 3 weeks

11/19/2016 12/06/2016 Game Playing 4 weeks

12/16/2016 01/13/2018 Comics Generation 4 weeks

參考文獻

相關文件

• User goal: Two tickets for “the witch” tomorrow 9:30 PM at regal meridian 16, Seattle. E2E Task-Completion Bot (TC-Bot) (Li et

 End-to-end reinforcement learning dialogue system (Li et al., 2017; Zhao and Eskenazi, 2016)?.  No specific goal, focus on

automatic pattern feature extraction from layers of neurons with backprop for GD/SGD. Lecture 3: Optimization in Deep Learning Difficulty of Deep

Most existing machine learning algorithms are designed by assuming that data can be easily accessed.. Therefore, the same data may be accessed

MR CLEAN: A Randomized Trial of Intra-arterial Treatment for Acute Ischemic Stroke. • Multicenter Randomized Clinical trial of Endovascular treatment for Acute ischemic stroke in

- Effective use of a variety of texts (e.g. information texts) to encourage deep reading, connect students’ learning experience and broaden their knowledge base. -

• Among the learning objectives for ELEKLA, the language development strategies, literary competence development strategies and attitudes specific to language and literature

• Among the learning objectives for ELEKLA, the language development strategies, literary competence development strategies and attitudes specific to language and literature