Jupyter notebooks from my blog posts at Puget Systems

• ML-LR-part1.ipynb -- Introduction to this series of posts about Linear Regression. I'm using this series as way to illustrate some of the key ideas of Machine Learning via a detailed analysis of Linear Regression.
• ML-LR-part2.ipynb -- Getting and Evaluating Data This notebook talks about pulling the King county house prices data-set from Kaggle and doing some visual analysis. To decide on a good data subset for the LR analysis.
• ML-LR-part3.ipynb -- Model and Cost Function an analysis and visualization of the linear model and cost function.
• ML-LR-part4.ipynb -- Parameter Optimization by Gradient Decent description, derivation, code and plots for gradient decent for linear regression.
• df_98039.csv -- King county house sales data for zipcode 98039, price and sqft_living.
• kc_house_data.csv -- King county house sales data, full.
• ML-LR-part5.ipynb -- Vectorization and Matrix Equations Derivation of the matrix equations including the gradient derivation and solution with the Moore-Penrose inverse.
• ML-LR-part6.ipynb -- Over/Under fitting and Non-Linear Feature Variables Last post in the linear regression series. Examples of using non-linear feature variables and the problem of over-fitting.
• ML-Logistic-Regression-theory.ipynb -- Logistic Regression Theory Discussion of the sigmoid function, how things work and derivation of equations including the gradient of the cost function in matrix form.
• ML-Logistic-Regression-Regularization.ipynb -- Logistic and Linear Regression Regularization Regularization to avoid over-fitting.
• CHDAGE.zip Data set of presence or absence of coronary heart disease by age. One of the Logistic Regression examples ...
• ML-Logistic-Regression-examples1 -- 2D data fit with multinomial model and 0 1 digits classification on MNIST dataset Also, some bug fixes to functions from the regularization notebook. Use these instead!
• **ML-Logistic-Regression-multinomial -- Logistic Regression: Multi-Class (Multinomial) -- Full MNIST digits classification example Multi-class classification with MNIST digits example