大數據分析, 機械學習, 資料採礦

Thu, 20 Aug 2020 00:00:00 +0000

關於大數據分析, 機械學習, 資料採礦等所使用的資料分析方法, 在統計領域也稱為統計學習, 統計主要是討論資料分析的方法, 因此統計學習的入們課程以討論資料分析為重心, 至於大數據資料庫的管理與存取, 不同演算法的差議, 不會在本課程討論. 另外, linear regression, logistic regression, data visualization 在統計基礎課程已經講授. machine learning 與 deep learning 可使用 Python, R, SAS, IBM-SPSS 等軟體.

Softwares

R:
- RSudio
- tidyverse
- ggplot2, lattice
- caret, tidymodels, mlr3
Python:
- Jupyter
- Jupyter notebooks
- NumPy, SciPy, Pandas
- Matplotlib, Seaborn
- Scikit-learn

大數據分析基本

Introduction
Tidy Data
Data Preprocessing
- Descriptive Statistics
- Missing Data
- Categorical Variables
- Dummy Variables
  - Combine Levels
  - Binning Predictors
- Continuous Variables
  - Near-zero Variance
  - Collinearity
  - Linear Dependencies
  - Outliers
- Data Transformations
- Putting It All Together

Big Data Visualization

Exploatory Data Analysis
R: Basic Graphics
R: lattice package
R: ggplot2 package
Analytic Visualization

Unsupervised Learning

PCA (principal Components Analysis)
Cluster Analysis
Basket Analysis and Association Analysis
Recommendation System

Supervised Learning

Predication: Theory and Evaluation
Classification: Theory and Evaluation
Linear Regression
Logistic Regression
Penalized Regression (L1, L2, Elastic Net)
Non-parametric Regression (LOWESS, Spline, GAM)
Robust Regression
Principal Component Regression (PCR)
Partial Least Square Regression (PLSR)
Discriminant (LDA, QDA)
Naive Bayes Classifier (NB)
k-Nearest Neighbor Classification (KNN)
Regression Tree and Classification Tree (Tree)
Support Vector Machines (SVM)
Bagging
Random Forest (RF)
Boosting

Text Mining

R: tidytext package
Word and Document Frequency
Relationships Between Words
Mining the Corpus
Sentiment Analysis
Topic Modeling

Deep Learning

Introduction to Neural Network
Recurrent Neural Network (RNN) and Elman Neural Networks (ENN)
Jordan Neural Networks (JNN)
Autoencoder
Stacked Autoencoder and Denoising Autoencoder (DA)
Restricted Boltzmann Machines (RBM)
Deep Belief Networks (DBN)
Convolutional Neural Network (CNN)
Natural Language Processing (NLP)

臨床試驗 Clinical Trials