Chapter 1 of machine learning practices

Chapter 1 of machine learning practices
Chapter 2 machine learning basics

Machine Learning Overview

Machine LearningIt is to convert unordered data into useful information.

We will use computers to demonstrate the true meaning behind the data.Machine Learning.

Machine Learning scenarios

For example, animal/CAT Pattern Recognition (official standard): People come to a conclusion through a large amount of experience to determine that it is a cat. Machine Learning (Data learning): people learn through reading and observe what it will call, small eyes, two ears, four legs, one tail, and draw a conclusion to determine that it is a cat. Deep Learning (deep data): people get to know about it and find that it is called "meow". It is similar to cats of the same type. Then, they come to the conclusion that it is a cat. Pattern recognition: pattern recognition is the oldest (as a term, it can be said to be outdated ). We collectively refer to the Environment and object as "patterns". recognition is a kind of cognition of patterns. It is how to let a computer program do something that looks "intelligent. By integrating wisdom and intuition, we can build programs to identify things, not people, such as recognizing numbers. Machine learning: machine learning is the most basic (one of the hot areas of startups and research laboratories ). At the beginning of 1990s, people began to realize that a method that can more effectively construct pattern recognition algorithms is to use data (which can be acquired through cheap labor) replace experts (people with a lot of image knowledge ). "Machine Learning" emphasizes that After inputting data to a computer program (or machine), it must do something to learn the data, the Learning steps are clear. Machine Learning is a specialized research on how computers simulate or implement human Learning behaviors to acquire new knowledge or skills, reorganizes existing knowledge structures to continuously improve their own performance. Deep learning: deep learning is a brand new and influential frontier. We don't even think about it-the post-deep learning era. Deep Learning is a new field in machine learning research. Its motivation lies in establishing and simulating a neural network for analysis and learning. It imitates the mechanisms of the human brain to interpret data, examples, sound and text. Reference address: http://www.csdn.net/article/2015-03-24/2824301http://baike.baidu.com/link? Url = 76P-uA4EBrC3G-I _ P1tqeO7eoDS709Kp4wYuHxc7GNkz_xn0NxuAtEohbpey7LUa2zUQLJxvIKUx4bnrEfOmsWLKbDmvG1PCoRkJisMTQka6-QReTrIxdYY3v93f55q

Machine Learning has been applied in many fields, far beyond the imagination of most people, spanning computer science, engineering technology, statistics, and other disciplines.

• Search engine: click your search to optimize your next search result.
• Spam: the spam ads are automatically filtered out to the bin.
• Supermarket coupons: you will find that when you buy a child diapers, the salesperson will give you a coupon for 6 cans of beer.
• Post Office Mail: handwritten software automatically identifies the address for sending greeting cards.
• Apply for a loan: make a comprehensive assessment based on your recent financial activities to determine whether you are qualified.

Machine learning components

• Category: divides instance data into appropriate categories.
• Regression: used to predict numeric data. (Example: fitting the optimal curve with a given data point)

Supervised Learning

• The value of the target variable must be determined so that the machine learning algorithm can discover the relationship between the feature and the target variable. (Including classification and regression)
• Sample Set: training data + Test Data
  • Training sample = feature + target variable (label: Classification-discrete value/regression-continuous value)
  • Features are usually columns in the training sample set, which are measured independently.
  • Target variable: The target variable is the test result of the machine learning prediction algorithm.
    • In classification algorithms, the type of the target variable is usually nominal (such as true and false), while in Regression Algorithms, the type of the target variable is usually continuous (such as: 1 ~ 100 ).
• Knowledge Representation:

Unsupervised learning

• The data has no category information and does not specify the target value.
• Clustering: in unsupervised learning, the process of dividing a dataset into multiple classes composed of similar objects is called clustering.
• Density Estimation: the process of finding and describing statistical values is called density estimation. The Probability Distribution of x is determined based on the training sample]
• In addition, unsupervised learning can reduce the dimensions of data features so that we can use 2D or 3D images to display data information more intuitively.

Training Process

Algorithm Summary

Machine Learning

Two considerations for algorithm selection

Example

Machine Learning Development Process

* Collect data: collect sample data * prepare data: note the data format * analyze data: to ensure that there is no spam data in the data set; skip this step if the algorithm can process data formats or trusted data sources. manual intervention is required to reduce the value of the automation system. * Training Algorithm: [machine learning algorithm core] If an unsupervised learning algorithm is used, skip this step because the target variable value does not exist. * Test Algorithm: [machine learning algorithm core] evaluating algorithm results * using algorithms: Converting machine learning algorithms into Applications

Python advantages

• Author: moment 1988
• GitHub address: https://github.com/apachecn/MachineLearning
• Copyright Disclaimer: Welcome to repost and learn => Please note that the information is from ApacheCN