Machine learning has been embedded in consumer products for decades, but we are seeing black box models in charge of prison sentencing, job recruiting, and mortgage approvals. We can trace the source of these bad models to biased data, practitioner ignorance, and a lack of good reproducible researching techniques.
This book will cover how to deal with bad data, techniques to make models that are easily interpreted by people and ways to mitigate the potential harm caused by the models. I aim to help those at the front lines of model building care more about how their models disproportionately impact people rather than tuning to increase the accuracy by .002%.
I'll show you how to persuade management to move projects in the direction that’s most considerate to the user.This book has been needed for decades because it not only helps the reader understand how human bias slips into models, but gives them code and techniques to analyze the models they’ve already built. This book will also give engineers the tools to push back on demands from management that result in harmful models.
This book is for anyone interested in Data Science, but is unsure where to start. Cut through the noise and learn my best tips for understanding Machine Learning with insight from my 4 years of industry experience. Learn the math as it applies to real-life data projects and get an understanding of fairness, ethics, and accountability in AI.
Abstract: Urban gun violence in cities across the world is a serious issue for public safety agencies and disaster management organizations. This led us to the development of the EDNA drone, an aerial robotics solution designed to equip first responders in high-risk settings with lifesaving-edge tools for situational awareness and non-lethal conflict resolution. The EDNA is an unmanned aerial vehicle (UAV) that delivers the patent-pending “Predictive Probable Cause” technology. The EDNA drone is designed to provide automated real-time analysis to assist teams entering high-risk situations where gun violence may occur.