Machine Learning - BKJackson/BKJackson_Wiki GitHub Wiki

Machine Learning Subpages:

AutoML and Pipelines

Introduction to Automated Machine Learning - Slides by Donald Whyte (Dec. 2016), Github
Simple Machine Learning Pipeline - Bringing together all essential parts to build a simple, but powerful Machine Learning pipeline. This will cover Keras/TensorFlow model training, testing, auto re-training, and REST API

Misc Machine Learning Links

Classifier calibration Useful when a classifier outputs probabilities.
10 more lessons learned from building real-life Machine Learning Systems, Xavier Amatriain, VP of Engineering, Quora at MLconf SF, 11/13/15
The future of machine learning at Pinterest Jan 2015
Open AI Gym Reinforcement learning.
Under the hood: PinQueue, a generic content review system "To date, PinQueue has helped us process more than three million items in more than 600 queues." March 11, 2106
Automatically Categorizing Yelp Businesses A multi-label classification problem.
How we increased active Pinners - By Optimizing Copy On the Copytune framework for optimizing copy.
Log Sum of Exponentials for Robust Sums Andrew Gelman. About overflow, underflow errors.
Memo Akten's ML Resources
R2D3 Visual Intro to Machine Learning Nice use of d3 graphics for illustrating decision trees.
Organizing an ML Project Portfolio Covers git setup, python, etc.
Time Series Prediction With Deep Learning in Keras Jason Brownlee
How Beginners Get It Wrong In Machine Learning J. Brownlee
Business Machine Learning - A curated list of practical business machine learning
Adaptive Basis Functions - Tutorial with python code examples

Basic curve fitting

Basic curve fitting with python

Least squares fitting with scipy.optimize

def model(theta, x):
    b, m = theta
    return m * x + b

def least_abs_deviation(theta, x, y):
    return np.sum(abs(model(theta, x) - y))

def least_squares(theta, x, y):
    return np.sum((model(theta, x) - y) ** 2)

from scipy.optimize import fmin
theta_guess = [0, 1]

theta_LAD = fmin(least_abs_deviation, theta_guess, args=(x, y))
theta_LS = fmin(least_squares, theta_guess, args=(x, y))

Engineering

API First for Data Science Pivotal Labs
Non-Mathematical Feature Engineering techniques for Data Science The biggest gains usually come from being smart about representing data, rather than using some sort of complex algorithm.