Takeaway: Andriy Burkov, head of Gartner’s Machine Learning team, recently released The latest edition of his Book The Hundred-Page Machine Learning Book, The goal is a machine learning book that anyone with basic math can read.
Eleven chapters in the latest edition of the book are available online, covering both supervised and unsupervised learning, including neural networks, deep learning, and some of the most important machine learning problems in computer science, math, and statistics. This book is easy to understand, suitable for beginners to learn and collect!
Big data official number back to the number “44”, check the download method to obtain the book.
Full text chapter Contents:
Part 01 Excerpt of preface
Let’s start with the truth: machines don’t learn. What a typical “Learning Machine” does is find a mathematical formula that, when applied to a set of inputs (called “training data”), produces the desired output.
This mathematical formula can also produce the correct output for most other inputs (unlike the training data), provided they come from the same or similar statistical distribution as the training data.
Why is this not learning? Because if you change or distort the input a little bit, the output is likely to be completely wrong. But that’s not how animals learn. If you’ve learned to play a video game by looking directly at the screen, you can still play well if someone turns the screen a little.
A machine learning algorithm, if it is trained by “looking directly at” the screen, will not be able to play a game on a rotating screen unless it is also trained to recognize rotation.
So why is it called machine learning? The reason is marketing: Arthur Samuel, an American pioneer in computer games and artificial intelligence, coined the term in 1959 while working at IBM. In a similar way to IBM’s attempt in 2010 to market the term “cognitive computing” as a way to distinguish itself from the competition, in the 1960s it used the new term “machine learning” to attract customers and talented employees.
As you can see, just as ARTIFICIAL intelligence is not intelligence, machine learning is not learning. Machine learning, however, is a universally accepted term that usually refers to the science and engineering of building machines that can do all sorts of useful things without explicit programming. Thus, the word “learning” in this term is used as an analogy for animal learning, not literally learning.
Content of part 02
Method of data Collection
Follow the public account [Pegasus Club]
Navigation recovery number [44]
You can view the download method
Previous welfare concerns about the pegasus public number, reply to the corresponding keywords package download learning materials; Reply “join the group”, join the Pegasus AI, big data, project manager learning group, and grow together with excellent people!