دانلود کتاب راه حل های کاربردی یادگیری ماشین با پایتون

The first computer program that can be categorized as a program that can learn, first came into picture in 1952, It was a game that played checkers, created by Arthur Samuel. He also coined the term “Machine Learning”. I wonder if he realized how popular this word will become. But the progress in the field of Artificial Intelligence halted till 1980s. With Hopfield’s recurrent neural networks, the neural networks started gaining researchers’ interest again. Then a groundbreaking idea of “Back propagation” was invented by Carnegie Mellon professor and computer scientist Geoffrey Hinton. With that, rapid advancements came into the field of machine learning and in 2006, he published a paper showing how to train a deep neural network capable of recognizing handwritten digits with state-of-theart precision (98%). He also coins the term “Deep Learning” and a revolution began. Fast forward to 14 years, and unless you are living under a rock, you have at least heard of machine learning. It now auto completes your sentences while you write your email, gives you movie suggestions, ranks your search queries, defeats professionals in their own game(OpenAI), raises security issues etc. It’s everywhere now. And it’s a highly attractive skill to know machine learning. As Arthur C. Clark, rightly said, “A sufficiently advanced technology is indistinguishable from magic”, Machine Learning embodies this statement better than any other technology in the current time. But the objective of this book is to break this spell. To bring machine learning into the stage and look deeper into its working. And by the time you finish reading this book, you will understand machine learning is not magic. It’s just sufficiently advanced technology with rapid development in computation hardware (cheap GPU, TPU, and so on.) and the availability of big data. These two are the cornerstones of modern deep learning. So, the objective of this book is to make machine learning more accessible and weaken the idea that a machine learning engineer needs to first develop statistics, linear algebra, machine learning concepts to become a machine learning practitioner. You need less of these and more experimentation and coding knowledge to do machine learning in the Industry. Research, however, is a field where a deeper understanding of underlying math and theory is necessary. But in my opinion, the programming aspect of machine learning is much more relevant than mere theoretical learning. Machine Learning is a new field and with any new field, you will get a lot of opinionated ideas. Things are not completely standardized in machine learning yet. Although some practices are common and provide good results almost all the time. In this book, you will find a lot of practices that are not common. These practices will provide you with tools to tackle any machine learning problems in Industry and provide you with techniques by which you can create your practices. This book also focuses on the practical implementation of machine learning ideas and play with real-world data and real-world problems. We will go through various data sources and understand the importance of data pipelines. We will go through a lot of examples and case studies that will give you an ample idea of how machine learning works in the Industry.