Toolbox
Preface
1 How we define Machine Learning
2 Preliminaries
I Formal Look at Prediction
Learning Systems
3 Bias-Variance Tradeoff
4 Overfitting
II Nonparametric Estimations
Parametric vs. Nonparametric methods
5 Parametric Estimations
- 5.1 Linear Probability Models (LPM)
- 5.2 Logistic Regression
6 Nonparametric Estimations - Basics
7 Smoothing
8 Nonparametric Classifier - kNN
III Self-Learning
9 Hyperparameter Tuning
10 Tuning in Classification
11 Classification Example
IV Tree-based Models
12 CART
13 Ensemble learning
14 Ensemble Applications
V SVM & Neural Networks
15 Support Vector Machines
16 Artificial Neural Networks
VI Penalized Regressions
Parametric models in prediction
17 Ridge
18 Lasso
19 Adaptive Lasso
20 Sparsity
VII Causality and Machine Learning
21 Model Selection
22 Sparsity
23 Use of Machine Learning in Causality
VIII Dimension Reduction Methods
Matrix Decompositions
24 Eigenvectors and eigenvalues
25 Singular Value Decomposition
26 Rank(r) Approximations
27 Moore-Penrose inverse
28 Principle Component Analysis
29 Factor Analysis
30 Dynamic Mode Decomposition
IX Time Series
Forecasting
31 ARIMA models
32 Grid search for ARIMA
33 Time Series Embedding
- 33.1 VAR for Recursive Forecasting
- 33.2 Embedding for Direct Forecast
34 Random Forest
35 Recurrent Neural Networks
X Network Analysis
Graphical Network Analysis
36 Fundementals
37 Regularized covariance matrix
XI Labs
38 R Lab 1 - Basics I
39 R Lab 2 - Basics II
- 39.1 Data frames and lists
- 39.2 Programming Basics
40 R Lab 3 - Preparing the data
- 40.1 Preparing the data for a regression analysis with lm()
- 40.2 “DUMMY” variable models
41 R Lab 4 - Simulation in R
XII Appendix
42 Algorithmic Optimization
43 Imbalanced Data
- 43.1 SMOTE
- 43.2 Fraud detection
References
© 2020 - 2022 Yigit Aydede

Toolbox for Social Scientists and Policy Analysts

Matrix Decompositions

Any matrix decomposition (and related topics) requires a solid understanding of eigenvalues and singular value decomposition (SVD).