lectures.md

| Home | Lectures | Labs | Assignments | Project| Contact |

Lecture Slides

ISLR = Introduction to Statistical Learning PDS = Python for Data Science (see homepage for links)

1. Week 1

• September 3
  • Intro to Data Science
  • Data Transformation
• September 5,
  • Tidy Data
  • Data Visualization
  • Tidy Intro, Tidy Advanced, and Vis notebooks

2. Week 2

• September 10
  • Function Fitting Intro
  • code examples
  • Reading: ISLR 2.1, 3.2.1, 3.5
• September 12
  • Function Fitting Algorithms
  • demo
  • Reading: ISLR 3.2.1, 3.3.1, 3.3.2, 3.5, 4.3, 7.2, 8.1

3. Week 3

• September 17
  • Function Fitting Implementation
  • Reading: PDS pg. 390 - 396, 400 - 405 and ISLR 4.6.6
  • Optional reading: ISLR 8.3.1, 8.3.2
  • Notebooks: Regression, knn + logistic regression, and trees
• September 19
  • Unsupervised Learning Algorithms
  • Reading: ISLR 10.1, 10.2, PDS pg. 462 - 476
  • Optional Reading: ISLR 10.3

4. Week 4

• September 24
  • Cross Validation + Model Selection
  • Reading: ISLR 5.1, PDS pg. 359 - 375
• September 26
  • Introduction to Inference
  • Reading: ISLR 3.1.1, 3.1.2, Bayesian Basics (intro - regression models)
  • Optional reading: MSMB 6.1 - 6.6, Statistics for Hackers

5. Week 5

• October 1
  • The Bootstrap
  • Reading: ISLR 5.2 - 5.3
  • Optional reading: CASI Chapter 1
• October 3
  • Large Scale Inference and Experimental Design
  • Reading: MSMB 6.7 - 6.11, 13.1 - 13.4.
  • Optional reading: CASI 15.1 - 15.3

6. Week 6

• October 8
  • Feature engineering + Outlier and error analysis
  • Reading: PDS 3
• October 10
  • Feature selection
  • the slides that are covered in the class are: slides 1 - 17 and 30 - 59
  • Reading: An Introduction to Variable and Feature Selection, KDD tutorial
  • Optional reading: Feature Selection for Data and Pattern Recognition, Computational Methods of Feature Selection, A Survey of Feature Selection Techniques
  • Feature selection tool scikit-learn, featureminer

7. Week 7 (Invited Guest Lectures)

• October 15
  • Geospatial and Time series data analysis by dr. Behrouz Babaki
  • Optional reading: Python Geospatial Development
  • Tools: Shapely, Basemap, geopandas
• October 17
  • Privacy and Transparency in Machine Learning by dr. Ulrich Aivodji
  • Optional reading: Composition Attacks and Auxiliary Information in DataPrivacy, A Survey Of Methods For Explaining Black Box Models, Privacy risk in machine learning- Analyzing the connection to overfitting, On the Protection of Private Information in Machine Learning Systems-Two Recent Approaches, fairwashing-the risk of rationaliztion
  • Protecting Privacy with MATH: Video

8. Week 8

• October 22
  • Reading week

9. Week 9

• October 31
  • MIDTERM: 16:30-18:30 at Z-110

10. Week 10

• November 5
  • Project presentation
• November 7
  • Algorithmic bias: Source and types of data bias, bias and discrimination in machine Learning: fairness metrics
  • Reading: Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries, Fairness definitions explained
  • Video: The Trouble with Bias, 21 fairness definitions and their politics
  • Fairness metric tools: The Aequitas Toolkit Paper, The Aequitas Tool

11. Week 11

• November 12
  • Text mining (NLP): part 1: NLP pipeline, BOW, n-grams, TF-IDF, and language model
  • Reading: Speech and Language Processing, Jurafsky and Martn, 2nd ed. (plus): Chapters 1 - 6
  • Tools: Python: NLTK, Gensim, Java: Stanford CoreNLP, Apache OpenNLP
• November 14
  • Text mining (NLP): part 2: Word embeddings, SVD, Word2Vec and GLoVe.
  • Reading: Word2Vec, GLoVe
  • Computer Vision: part 1: CV pipleline, Warping, Point processing, Filters
  • Reading: Computer Vision:Algorithms and Applications, Richard Szeliski, Chapters 3.1.1, 3.2, 3.6.1
  • Tools: Python: OpenCV

12. Week 12

• November 19
  • Computer Vision: part 2: Convolution
  • Reading: Computer Vision:Algorithms and Applications, Richard Szeliski, Chapters 3.1.1, 3.2, 3.6.1, Filtering, Convolution, Correlation, Chapter 5
• November 21
  • Crash course to deep learning: Perceptrons, Neural networks, Convolution Neural networks, Recurrent Neural Network and LSTM
  • Reading: Deep Learning, Ian Goodfellow, Yoshua Bengio, Aaron Courville, Chapter 6, 9, 10
  • ConvNet notes
  • Tools: Tensorflow, Keras, Pytorch

13. Week 13

• November 26
  • Graph ML: part 1: node embeddings: adjajency matrix,matrix factorization, multi-hop embedding, random walk embeddings, and node2vec
  • Reading: Representation learning on graphs
  • Tools: Python: igraph, NetworkX
  • Tool: embedding tools: node2vec, deepwalk, GraphVite
• November 28
  • Graph ML: part 2: Graph neural networks, and Graph convolutional networks
  • Multimodal Learning: Multimodal machine learning, Multimodal Representation, Fusion
  • Tools: Deep autoencoders in Keras, Ensemble methods: scikit-learn
  • Reading: Multimodal Machine Learning: A Survey and Taxonomy

14. Week 14

• December 3
  • Final project presentation

15. Week 15

• December 10
  • FINAL: 16:30-18:59 at Z310 and Z330