Managing Machine Learning Projects

Abstract

Anyone who’s written a book knows it’s an unreasonably hard thing to do. I’ve needed a lot of help. Doug Rudder, my editor, and the team at Manning exceeded expectations and helped me transform a huge random mess of a manuscript into something I hope is much more useful to readers.. I don’t think that anyone who hasn’t worked with Manning can really know just how much value they add. This book could be a lot better if someone else wrote it, but without the work that everyone at Manning put in, it would be immeasurably worse. Manning arranged an extensive reviewing process that provided me with ano- nymized feedback, of course, I don’t know who did which review, but every review was immense: Andrei Paleyes, Chris Fry, Darrin Bishop, Florian Roscheck, Igor Vieira, João Dinis Ferreira, Kay Engelhardt, Khai Win, Kumar Abhishek, Lakshminarayanan AS, Laurens Meulman, Maria Ana, Marvin Schwarze, Mattia Di Gangi, Maxim Volgin, Ricardo Di Pasquale, Richard Dze, Richard Vaughan, Sanket Naik, Sriram Macharla, Vatsal Desai, Vojta Tuma, William Jamir Silva. The amount of work, attention to detail and honest, direct input that you provided was just amazing. Thank you, if and when we meet up collar me for a beer or beverage of your choice. I owe you one for sure. I have been very fortunate to have some amazing mentors in my career, and one of the most important things I think that anyone can do is to find some people who will help you as you develop your skills and abilities. Professor Max Bramer gave me an amazing start in machine learning when he took me on as a PhD student, I had four brilliant years of exploring everything that ML could offer in the mid-1990s, and that changed my life. Paul O’Brien took a similar risk when he recruited me at BT Labs, Paul is my pro- fessional role-model, the manager and mentor I aspire to be. Literally, whenever I have a problem at work I think “what would Paul do”. The other thing that everyone needs is colleagues who will indulge your ideas and peculiar thinking, point out where you are wrong, and share their own thoughts. For this I would like to particularly thank Rob Claxton who spent hundreds of hours talking to me on any and every topic to do with Data Science, AI and ML. There were many other people at BT, The Turing Institute, and MIT who were prepared to let me test their patience and gave me time I didn’t deserve, but the conversations I’ve had with Rob over the last twenty odd years were (and are) intellectually formative for me. When I was writing this book, I was generally bad-tempered, preoccupied, and generally insufferable. My wife, Buffy, and my daughter, Arwen, put up with this non- sense sometimes, but mostly told me to stop it. Which was what I needed. Buffy and Arwen, I love you very much. Thank you everyone