查詢教學大綱與進度

教學大綱與進度

課程基本資料：

學年期課號課程名稱階段學分時數修教師班級人撤備註

111-1 306347 高等類神經網路應用 1 3.0 3 ★ 田方治工管所 13 0 　

學年期	課號	課程名稱	階段	學分	時數	修	教師	班級	人	撤	備註
111-1	306347	高等類神經網路應用	1	3.0	3	★	田方治	工管所	13	0

教學大綱與進度：

教師姓名田方治　教師諮商時間(Office Hours)

Email fctien@ntut.edu.tw

最後更新時間 2022-06-22 09:05:57

課程大綱本課程之主要目的除介紹基本之類神經理論外，更在於運用現有之類神經網路於工業工程領域內之相關運用，將以現有之資料庫實地解決分類、分群、預測、監控等實際問題，使學生能夠了解類神經網路之基礎理論外，更可深入運用此工具。 The objective of this class is to explore the basic theoretical basis and applications of Neural Networks and their applications. In this course, students will learn how to build deep learning networks and their applications in machine vision areas 加強程式、邏輯訓練及實作 (PyTorch & python & MVA – prerequisites) The major topics of networks covered include: AD(Anomaly Detection): CS-Flow、PatchCore、Padim Classifier: CNN (ResNet, DenseNet, EfficientNetv2) OD (Object Detection): FRCNN、YOLO、Mask RCNN Semantic Segmentation: AutoEncoder, UNet GAN: DCGAN etc.

課程進度 The content of this course include the following subjects： 1. PyTorch: (3 weeks) a. Basic commands & nn.module related commands b. Gradient descent & BPN c. CNN & LeNet d. Dataset & Dataloader e. Loss Function & Transformation 2. CNN Training Tips & CAM Heatmaps, Visualization (2 week) 3. Classifier: ResNet, DenseNet, EfficientNet (2 weeks) 4. OD (Object Detection): FRCNN, Yolo, Mask RCNN, Labeling (2 weeks) 5. AD (Anomaly Detection): PaDim, CS-Flow, PatchCore (2 weeks) 6. Semantic Segmentation: AE, Unet (2 Weeks) 7. GAN: DCGAN, etc. (2 Weeks)

評量方式與標準 Assignment & Lab 35 % Final Test 25% Paper Study and Presentation 10 % Term Project - Competition 30 % AC Book Reading 5% (optional)

使用教材、參考書目或其他 【遵守智慧財產權觀念，請使用正版教科書，不得使用非法影印教科書】

使用外文原文書：否
Satish Kumar, Neural Networks— a classroom approach, McGrawHill, 滄海書局 (Textbook) Simon, Haykin, Neural Network and Learning Machine, Pearson International Youtubes: 田方治: MVA、Deep Learning Others Personal power points

課程諮詢管道 Google meet 上課連結: https://meet.google.com/pwu-amgf-sgm

備註
1. 線上上課
2. 連結: https://meet.google.com/pwu-amgf-sgm

教師姓名	田方治　教師諮商時間(Office Hours)
Email	fctien@ntut.edu.tw
最後更新時間	2022-06-22 09:05:57
課程大綱	本課程之主要目的除介紹基本之類神經理論外，更在於運用現有之類神經網路於工業工程領域內之相關運用，將以現有之資料庫實地解決分類、分群、預測、監控等實際問題，使學生能夠了解類神經網路之基礎理論外，更可深入運用此工具。 The objective of this class is to explore the basic theoretical basis and applications of Neural Networks and their applications. In this course, students will learn how to build deep learning networks and their applications in machine vision areas 加強程式、邏輯訓練及實作 (PyTorch & python & MVA – prerequisites) The major topics of networks covered include: AD(Anomaly Detection): CS-Flow、PatchCore、Padim Classifier: CNN (ResNet, DenseNet, EfficientNetv2) OD (Object Detection): FRCNN、YOLO、Mask RCNN Semantic Segmentation: AutoEncoder, UNet GAN: DCGAN etc.
課程進度	The content of this course include the following subjects： 1. PyTorch: (3 weeks) a. Basic commands & nn.module related commands b. Gradient descent & BPN c. CNN & LeNet d. Dataset & Dataloader e. Loss Function & Transformation 2. CNN Training Tips & CAM Heatmaps, Visualization (2 week) 3. Classifier: ResNet, DenseNet, EfficientNet (2 weeks) 4. OD (Object Detection): FRCNN, Yolo, Mask RCNN, Labeling (2 weeks) 5. AD (Anomaly Detection): PaDim, CS-Flow, PatchCore (2 weeks) 6. Semantic Segmentation: AE, Unet (2 Weeks) 7. GAN: DCGAN, etc. (2 Weeks)
評量方式與標準	Assignment & Lab 35 % Final Test 25% Paper Study and Presentation 10 % Term Project - Competition 30 % AC Book Reading 5% (optional)
使用教材、參考書目或其他	【遵守智慧財產權觀念，請使用正版教科書，不得使用非法影印教科書】使用外文原文書：否 Satish Kumar, Neural Networks— a classroom approach, McGrawHill, 滄海書局 (Textbook) Simon, Haykin, Neural Network and Learning Machine, Pearson International Youtubes: 田方治: MVA、Deep Learning Others Personal power points
課程諮詢管道	Google meet 上課連結: https://meet.google.com/pwu-amgf-sgm
備註	1. 線上上課 2. 連結: https://meet.google.com/pwu-amgf-sgm