ffund / ml-notebooks

Extra notebooks for ECE-GY 6143

Machine learning notebooks

These are notebooks developed for ECE-GY 6143 Intro to Machine Learning at NYU Tandon School of Engineering.

Intro ML, exploratory data analysis

• Notebook: Python + numpy tutorial
• Notebook: Colab tutorial
• Notebook: Printing from Colab
• Notebook: Exploratory data analysis (in-class)
• Notebook: Exploratory data analysis (homework)
• Notebook: Data detective challenge (optional homework)

Linear regression

• Notebook: Linear regression in depth
• Notebook: Compute regression coefficients by hand
• Notebook: Regression metrics
• Notebook: Case study on "Beauty in the Classroom"
• Notebook: Residual analysis on Advertising data (homework)

Gradient descent

• Notebook: Gradient descent in depth

Bias-variance tradeoff, model selection

• Notebook: Bias-variance tradeoff and model selection in depth
• Notebook: Model order selection for neural data (homework)

Regularization

• Notebook: Regularization in depth

Logistic regression

• Notebook: Logistic regression in depth
• Notebook: Logistic regression for handwritten digits classification
• Notebook: COMPAS case study
• Notebook: Classifying your own handwritten digit (homework)

K nearest neighbor

• Notebook: K nearest neighbor in depth
• Notebook: Voter classification with K nearest neighbor (homework)

Decision tree, ensemble methods

• Notebook: Decision trees and ensembles
• Notebook: AdaBoost
• Notebook: Bias and variance of KNN and decision tree models

Support vector machines

• Notebook: Support vector machines
• Notebook: Handwritten digits classification
• Notebook: Bias and variance of SVM
• Homework: Grid search for SVM hyperparameter tuning (use NYU Google account to open)

Neural networks

• Notebook: Backpropagation from scratch
• Notebook: Draw your own classification problem for a neural network
• Homework: Neural network for musical instrument classification (homework) [

Convolutional neural networks

• Deep dive: Convolutional neural networks (use NYU Google account to open)
• Transfer learning (use NYU Google account to open)
• Homework: Transfer learning

Unsupervised learning

Project topics

For your project, you will replicate and then extend an existing ML project (typically a recent publication in a major ML conference). See projects list for examples.