校際選修

115-1 選課時程

進行中

  • 初選第一階段 6/15/2026
  • 初選第二階段 6/22/2026
  • 校際選修 8/24/2026
  • 初選第三階段 8/31/2026
  • 開學後加退選 9/7/2026
  • 逾期加退選 9/21/2026
選課資源

深度生成模型

Deep Generative Models

學期
114-2
學分
3 學分
當期課號
639002
永久課號
AICA30011
開課單位
智慧科學暨綠能學院
授課教師
魏澤人
校區
歸仁
類別
選修
上課時間表
週三
2
09:00–09:50
深度生成模型
CM218
3 節連堂
3
10:10–11:00
4
11:10–12:00

* 根據陽明交大上課時間表所列

概述

Generative models are a method in AI and machine learning that are widely used in various fields, in particular, generative AI such as large language models. They can serve as a foundational approach, used in conjunction with other algorithms, for example, as part of data augmentation, or they can simply be used to generate more content. In recent years, there have been many advances in generative models due to the development of deep learning, such as Diffusion models, GANs, VAEs, autoregression models, Generative flow, and so on. In this course, we will start with an introduction to these foundational models and then proceed to cover some of the latest developments.

先修科目

Deep learning, linear algebra

教學方式

course website https://tjwei.tw/

評分方式

homework 100%

週次計畫
週次主題
第 1 週Introduction
第 2 週Introduction: Autoencoder
第 3 週Introduction: Autoregressive model
第 4 週Introduction: n-gram and language generalization
第 5 週VAE: Variational Autoencoder
第 6 週GAN: GAN, DCGAN and math
第 7 週GAN: WGAN and SNGAN
第 8 週Conditional GAN
第 9 週Pix2Pix and CycleGAN
第 10 週Diffusion Model, Generative Flow
第 11 週Diffusion Model. Theory and Stochastic Differential Equation
第 12 週Diffusion Model and Language Model
第 13 週Generative Language models: Applications and Theory
第 14 週Generative Language models: Implementations
第 15 週Recent advances
第 16 週Summary and review
教科書

online