#
Notes On Using

Python and Data Science

To Make Cool Stuff

For a good while, whenever I'd start to learn something new, I'd go down a rabbit hole of documentation, books, YouTube videos, etc. As I dug deeper and deeper into the material, I'd leave behind mountain of scratch paper where I'd jotted along. Then (if I even found the time) I'd rewrite/distill it into some Moleskine notebook for "easy reference" later.

Then, of course, you have to know how to reference all of your reference materials. If I forgot to bring my notebook to work, I'd be out of luck. Or better yet, how would I share the notes that I took if they might be useful for someone else?

So, by the time I found Chris Albon's excellent note-compiling repository it felt like the answer to all of my problems. After a good deal of reverse-engineering, this page is less about curating a body of work, and more a consistent place that I can stage the things I'm figuring out. That means that I don't intend to brain-dump all of the things I know just to fill space. **If you're new** and looking for beginner materials, this might prove more helpful!

### Python

#### Object Oriented Programming

### Machine Learning

#### Regression

- Evaluation using Residual Plots
- F Statistic
- Interaction Terms in Python
- Leverage, Influence, and Cook's Distance
- Logistic Regression Basics
- Logistic Regression Gradient Descent
- Multi-Class Regression with SoftMax
- Partial Least Squares
- R Squared Measure of Fit
- Ridge and Lasso Regression
- Splines and Generalized Additive Models

#### Tree-Based Models

#### Unsupervised Methods

#### Neural Networks: Basics

- Activation Functions
- Back Propagation
- Batch Normalization
- Bias, Variance, and Regularization
- Forward Propagation
- Forward and Back Prop in Deeper Networks
- Mini Batch Gradient Descent
- Model Improvement Strategy Heuristics
- Momentum, RMSprop, and Adam Optimization for Gradient Descent
- Representing XNOR via a simple Net

#### Neural Networks: Computer Vision

- CNN organization
- Convolution Hyperparameters
- Convolution Intuition
- Fine-Tuning Pretrained Networks
- Image Data Augmentation
- Inception Architecture
- Interpretability: Decision Attribution
- Interpretability: Find the Essence of Filters
- Interpretability: Visualizing Intermediate Activations
- Neural Style Transfer
- Object Detection Rough Intuition
- Pooling
- Residual Networks
- VGG Architecture