دانلود کتاب یادگیری ماشین خودکار با AutoKeras

Can deep learning be accessible to everyone? Without a doubt, this is the objective that the cloud services offered by giants such as Google or Amazon are trying to achieve. Google AutoML and Amazon ML services are cloud-based services that make it easy for developers of all skill levels to use machine learning technology. AutoKeras is the free open source alternative and, as we’ll see soon, a fantastic framework. When faced with a deep learning problem, the choice of an architecture or the configuration of certain parameters when creating a model usually comes from the intuition of the data scientist, based on years of study and experience. In my case, being a software engineer without a broad background in data science, I have always looked for methods to automate this part, using different search algorithms (grid, evolutionary, or Bayesian) to explore the different variables that make up a model. Like many other Python developers, I started in the world of machine learning with scikit-learn and then jumped into deep learning projects with TensorFlow and Keras, testing different frameworks such as Hyperas or TPOT to automate model generation and even developed one to explore architectures in my Keras models, but once AutoKeras was released I found everything I needed, and since then I’ve been using it and contributing to the project. AutoKeras has a large community that grows day by day and is supported by the widely known deep learning framework Keras, but apart from its documentation and the occasional blog article, to date, there are almost no books written about it– this book tries to fill that gap. Both the book and the framework, are aimed at a broad spectrum of ML professionals, from beginners looking for an alternative to cloud services (using it as a black box simply by defining its inputs and outputs), to seasoned data scientists who want to automate exploration by defining search space parameters in detail and exporting generated models to Keras for manual fine tuning. If you are one of the first, maybe these terms and concepts may sound strange to you, but do not worry, we will explain them in detail throughout the book.