國立中興大學教學大綱

課程名稱	(中) 數據科學方法(5139)
	(Eng.) Fundamental Methodology for Data Science
開課單位	資科所
課程類別	選修			學分	3	授課教師	蔡鴻旭
選課單位	應數系 / 學士班			授課使用語言	中文	英文/EMI		開課學期	1141
課程簡述	This Data Science Course is introductory to Data Science. The syllabus of the course will introduce students to methods of processing data before dealing data, and concepts of data science algorithms for machine learning that help to gain some meaningful insights from structure or unstructured data. Statistics courses just analyze the history of the data, but with the help of data science courses and machine learning algorithms, students can predict future trends (profit, loss, or other insights). Therefore, the prediction results help students who want to learn how to make data-driven decisions and analyze ways to maximize the profit of an organization using data. The course will introduce the machine learning and/or deep learning algorithms for artificial intelligent approaches of learning from data.
先修課程名稱								課程含自主學習	Y

課程與核心能力關聯配比(%)			課程目標之教學方法與評量方法
課程目標	核心能力	配比(%)	教學方法	評量方法
Let students understand basic math background for data science. Let students understand approaches of learning from data. Let students understand the trend on the development of data science. Let students develop data models for real problems.	1.數理基礎知識 4.計算科學專業知識	50 50	專題探討/製作 討論 講授	書面報告 出席狀況 口頭報告 作業 實作
授課內容(單元名稱與內容、習作/每週授課、考試進度-共16週加自主學習)
週次 授課內容 第1週 第1週: 說明課程大綱、簡述數據(資料)科學、說明實作環境、電腦作業、專題計畫、論文研讀、加分(Week #1: Introduction of syllabus, Data Science, building programming environment for data science, Computer assignment, Computer Project, Journal paper study and presentation,人工智慧共創平台，Aidea, 興大農業資料共享平台, https://aidata.nchu.edu.tw) 分組(每組1~2位同學) (1~2 classmates for each team) 第1週: 人工智慧與大數據概述(Week #1: Introduction of Artificial Intelligence and Big Data) 第2週 第2週:資料與資料準備(Week #2: Data and data preparation) 第2-4週: 機器學習概述及應用實作(Week #2-#4: Introduction of machine learning and its applications) 機器學習應用 (Applications of Machine Learning) 關聯法則(Association rule) 單純貝氏分類器(Naïve Bayes classifier) 羅吉斯迴歸(Logistic Regression) 第3週 第2-4週: 機器學習概述及應用實作(Week #2-#4: Introduction of machine learning and its applications) 決策樹(Decision Trees) 隨機森林(Random Forests) K近鄰演算法 (K Nearest Neighbors algorithm) 第4週 第2-4週: 機器學習概述及應用實作(Week #2-#4: Introduction of machine learning and its applications) 降維(Dimensionality reduction) 分群(Clustering) 第5週 第5-6週:類神經網路概述及應用實作(Week #5-#6: Introduction to Neural network and its application) 類神經網路簡介(Introduction to Neural network) 感知元件之基本結構(Architecture of Perceptrons) 感知元件之多層結構(Multi-layers architecture of Perceptrons) 感知元件之矩陣計算模式(Matrix computation for Perceptrons) 感知網路之認知功能(Recognition for Perceptron Network) 認知學習演算法(Learning algorithm for Perceptron Network) 類神經網路應用-電腦進行手寫辨識(Application on artificial Neural network - Handwriting image recognition) 第6週 第5-6週:類神經網路概述及應用實作(Week #5-#6:Introduction to Neural network and its application) 感知器 (Perceptron) 多層感知器 (Multi-layers Perceptron) 支持向量機 (Support Vector Machine) 支持向量迴歸 (Support Vector Regression)  back-propagation(BP) gradient-vainshing regulation 第7週 第7週:深度學習-卷積類神經網路來建模應用於分類模型設計與實作(Week #7:Deep Learning- using Convolutional Neural Network (CNN) on the design of classifications and practical exercises) 深度學習簡介(Introduction to deep learning) 深度學習-CNN 模型 (deep learning for CNN) CNN實作介紹(practical exercises for CNN) Keras CNN 辨識手寫數字 (using Keras CNN on recognition for MNIST handwritten digit database) Keras Cifar-10影像辨識資料集(Keras Cifar-10 image database) Keras CNN 辨識Cifar-10影像(Using Keras CNN on recognition for Cifar-10 image database) 第8週 第8週: 基於CNN的模型介紹 (Week #8: Models based on CNN) LeNet VGGNet Residual Network DenseNet U-Net InceptionNet(GoogLeNet) Fully Convolutional Networks(FCNs) MobileNet V1 EfficientNet MaskRCNN Capnet VAE GAN 第9週 第09-10週 期中報告-電腦作業報告、電腦專題精簡計畫書報告&論文研讀報告(Week #10-#11: midterm report) 電腦作業報告-各組報告 (ppt) (Computer project #1 report and presentation for each team) 電腦專題精簡計畫書(word)&報告(ppt)-各組 (Draft of project report with word and presentation by ppt for each team) 論文研讀報告(ppt)(第一篇論文)-每位同學(Present 1st paper by ppt for each student) 第10週 第09--10週 期中報告-電腦作業報告、電腦專題精簡計畫書報告&論文研讀報告(Week #10-#11: midterm report) 電腦作業報告-各組報告 (ppt) (Computer project #1 report and presentation for each team) 電腦專題精簡計畫書(word)&報告(ppt)-各組同學 (Draft of project report with word and presentation by ppt for each team) 論文研讀報告(ppt)(第一篇論文)-每位同學 (Present 1st paper by ppt for each student) 第11週 第11週: 卷積類神經網路應用專題-影像分類、影像檢索、影像分割 (Week #9:CNN applications for image classification, image retrieval and image segmentation) Autoencoder and decorder (image transformer) Transfer Learning You Only Look Once(YoLo) RCNN/Fast-RCNN Multi-label/Multiclass/Multilabel Multitask 臉手識別實現答題系統 (face-and-hand recognition for interactive response system) 表面瑕疵偵測技術(surface defect detection) 應用深度學習研製智慧行動辨識不同災害類型影像(image classification for different damages on images for soil-and-water conservation) 基於深度學習發展智慧型影像實例分割與分類應用於水土保持影像分析之研究(image classification and instance segmentation for images for soil-and-water conservation) 第12週 第12週: (Week #12: paper survey for CNN applications) -Content-Based Image Retrieval Based on CNN and SVM --VAE-Y-Autoencoders-disentangling latent representations via sequential-encoding -Neural network-based multi-task learning for inpatient flow classification and length of stay prediction -DML-PL Deep metric learning based pseudo-labeling framework -Hard Sample Aware Noise Robust Learning 第13週 第13週: 深度學習-長短期記憶模型來建模應用於預測模型設計與實作(Week #13: deep learning - Long short term memory model for the design of prediction models and practical exercises) 深度學習-遞歸神經 網路-長短期記憶模型 (Deep Learning–RNN-LSTM) (Note: 遞歸神經網路，recurrent neural networks(RNN) and 長短期記憶模型 Long short-term memory，LSTM) 第14週 第14-16週 期末報告-電腦作業報告&論文研讀報告(Week #14-#16: final-term report) 電腦作業報告書(word)&報告(ppt)-各組(computer project #2 report with word format and presentation by ppt for each team) 論文研讀報告(ppt)(第二篇論文) (Present 2nd paper by ppt for each student) 第15週 第14-16週 期末報告-電腦作業報告&論文研讀報告(Week #14-#16: final-term report) 電腦作業報告書(word)&報告(ppt)-各組 (computer project #2 report with word format and presentation by ppt for each team) 論文研讀報告(ppt)(第二篇論文) (Present 2nd paper by ppt for each student) 第16週 第14-16週 期末報告-電腦作業報告&電腦專題精簡成果報告書(Week #14-#16: final-term report) 電腦作業報告書(word)&報告(ppt)-各組 (computer project #2 report with word format and presentation by ppt for each team) 電腦專題成果報告書(word)&報告(ppt)-各組 (final project report with word format and presentation by ppt for each team) 第17-18週 自主學習 (Week #17-#18: self-learning) 至少15篇sci/ssci期刊論文(a report on at least 15 sci/ssci journal papers for study for each student) 第17-18週 自主學習 (Week #17-#18: self-learning) 至少15篇sci/ssci期刊論文(a report on at least 15 sci/ssci journal papers for study for each student) 自主學習 內容
學習評量方式
Participation: 15%; Paper presentation: 20%; Computer assignment and report: 15%; Project (Project proposal, result and report): 35%; self-learning report 15%; Bonus: 20%
教科書＆參考書目(書名、作者、書局、代理商、說明)
==Textbook 大數據分析與資料挖礦 2/e，作者： 簡禎富 、 許嘉裕，(前程文化) Big-data-analytics-Data-mining 2/e, authors: Hsu, Chia-yu and Chien, Chen-fu (https://www.fcmc.com.tw/) 製造數據科學，作者：李家岩、洪佑鑫 (前程文化) Data Science in Manufacturing, authors: Hung, Yu-Hsin Jeff and Lee, Chia-Yen (https://www.fcmc.com.tw/) ===References== Data Science from Scratch: First Principles with Python, 2/e (O’Reilly) Data Science from Scratch｜用 Python 學資料科學, 2/e (中文版) (碁峰資訊) https://github.com/joelgrus/data-science-from-scratch ====== Géron, Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow (2nd ed.) 書刊名: 精通機器學習使用Scikit-Learn, Keras與TensorFlow，作者: 杰龍 (Géron, Aurélien)，其他作者: 賴屹民，(歐禮萊) ====== https://scikit-learn.org/stable/user_guide.html === Lau et al., Principles and Techniques of Data Science. https://ds100.org/sp18/assets/lectures/lec01/01-intro-to-data100_v2.pdf == Python資料科學與機器學習：從入門到實作必備攻略 (博碩文化) ===== Raschka, Sebastian, and Vahid Mirjalili. Python Machine Learning, 3rd Ed. Packt Publishing, 2019. https://github.com/rasbt/python-machine-learning-book-3rd-edition Python機器學習第三版（上）譯者：劉立民、吳建華　譯(博碩文化) Python機器學習第三版（下）譯者：劉立民、吳建華　譯(博碩文化) ==== 資料科學的建模基礎 - 別急著coding！你知道模型的陷阱嗎？作者：江崎貴裕 著、王心薇 譯、施威銘研究室 監修(旗標) === Deep Learning, Ian Goodfellow, Yoshua Bengio, Aaron Courville, MIT Neural Networks: A Comprehensive Foundation, Simon Haykin http://airobot.ccu.edu.tw/chapter-2-%e5%be%9e%e6%84%9f%e7%9f%a5%e7%b6%b2%e8%b7%af%e8%aa%aa%e8%b5%b7/ == 深度學習-影像處理應用，彭彥璁、李偉華、陳彥蓉，全華圖書 (Deep learning for image processing applications), 2023.06.
課程教材（教師個人網址請列在本校內之網址）
http://www.amath.nchu.edu.tw/member_detail.php?Key=71
課程輔導時間
星期二 Tuesday 8th-9th classes
聯合國全球永續發展目標(連結網址)
04.教育品質 提供體驗課程：N

請尊重智慧財產權及性別平等意識，不得非法影印他人著作。
更新日期 西元年/月/日：無	列印日期 西元年/月/日：2025 / 7 / 02
MyTB教科書訂購平台：http://www.mytb.com.tw/