News classification
Contact us
- Add: No. 9, North Fourth Ring Road, Haidian District, Beijing. It mainly includes face recognition, living detection, ID card recognition, bank card recognition, business card recognition, license plate recognition, OCR recognition, and intelligent recognition technology.
- Tel: 13146317170 廖经理
- Fax:
- Email: 398017534@qq.com
What exactly is machine learning?
What exactly is machine learning?
With the development of computer technology, we now have the ability to store and process large amounts of data and access data from remote sites via computer networks. At present, most data access devices are digital devices, and the recorded data are reliable. Take a supermarket chain, for example, with hundreds of stores across the country, and several thousand items of retail services for millions of customers. The point of sale equipment records the details of each transaction, including dates, customer identification numbers, purchases of goods and quantities, total consumption, etc.. This is typical of several G bytes of data per day. These data can become useful only when analyzing the data and converting it into useful information, such as making predictions.
machine learning
We don't know exactly what people are inclined to buy or which author to recommend to someone who likes to read Hemingway. If we know, we don't need any data analysis; we'll just supply and record the coding. But just because we don't know, we can only collect data and expect to extract answers to these or similar questions from the data.
Gl hwein is a kind of warm, a little sweet and spicy Wine. Christmas is very popular in europe. The translator is sure that there is a process that can explain what we have observed. Although we don't know the details of the data generation process (for example, customer behavior), we know that data creation is not entirely random. People don't go to supermarkets to buy goods at random. When people buy beer, will buy chips; summer and winter to buy ice cream, Gl hwein buy perfume?. Certain patterns exist in the data.
We may not be able to fully identify the process, but we believe that we can construct a good and useful approximation (good, and, useful, approximation). Although such an approximation is not yet possible to explain everything, it can still explain some parts of the data. We believe that although identifying the whole process may be impossible, we can still find some patterns or regularities. This is the location of machine learning. These models can help us understand the process, or we can predict the use of these models: suppose that in the future, at least not far in the future, and will not collect the sample data is greatly different, the future will also be expected to forecast is correct.
The application of machine learning methods in large databases is called data mining (data mining). A similar situation, such as a large amount of metal oxides and raw materials mined from a mine, which produces a very small amount of very precious material. Similarly, in data mining, large amounts of data need to be processed to build simple and useful models, such as predictive models with high accuracy. The application of data mining is widely: in addition to outside the retail industry, in the financial industry, the analysis of historical data of their bank, for the construction of the application model of credit analysis, fraud detection, stock market and so on; in the manufacturing industry, the learning model can be used for optimization, control and fault detection; in the field of medicine, can be used for learning program medical diagnosis; in the field of telecommunications, communication mode analysis can be used for network optimization and improve the service quality; in the field of scientific research, such as physics, biology and astronomy data only by computer analysis may be fast enough. The world wide web (World Wide Web) is vast and growing, so retrieval of relevant information on the World Wide Web cannot rely on manual completion.
Machine learning uses instance data or past experience to train a computer to optimize some performance criteria. We have models that depend on certain parameters, and learning is the execution of a computer program that uses training data or previous experience to optimize the parameters of the model. Models can be predictive (predictive) for future forecasts or descriptive (descriptive),
machine learning
We don't know exactly what people are inclined to buy or which author to recommend to someone who likes to read Hemingway. If we know, we don't need any data analysis; we'll just supply and record the coding. But just because we don't know, we can only collect data and expect to extract answers to these or similar questions from the data.
Gl hwein is a kind of warm, a little sweet and spicy Wine. Christmas is very popular in europe. The translator is sure that there is a process that can explain what we have observed. Although we don't know the details of the data generation process (for example, customer behavior), we know that data creation is not entirely random. People don't go to supermarkets to buy goods at random. When people buy beer, will buy chips; summer and winter to buy ice cream, Gl hwein buy perfume?. Certain patterns exist in the data.
We may not be able to fully identify the process, but we believe that we can construct a good and useful approximation (good, and, useful, approximation). Although such an approximation is not yet possible to explain everything, it can still explain some parts of the data. We believe that although identifying the whole process may be impossible, we can still find some patterns or regularities. This is the location of machine learning. These models can help us understand the process, or we can predict the use of these models: suppose that in the future, at least not far in the future, and will not collect the sample data is greatly different, the future will also be expected to forecast is correct.
The application of machine learning methods in large databases is called data mining (data mining). A similar situation, such as a large amount of metal oxides and raw materials mined from a mine, which produces a very small amount of very precious material. Similarly, in data mining, large amounts of data need to be processed to build simple and useful models, such as predictive models with high accuracy. The application of data mining is widely: in addition to outside the retail industry, in the financial industry, the analysis of historical data of their bank, for the construction of the application model of credit analysis, fraud detection, stock market and so on; in the manufacturing industry, the learning model can be used for optimization, control and fault detection; in the field of medicine, can be used for learning program medical diagnosis; in the field of telecommunications, communication mode analysis can be used for network optimization and improve the service quality; in the field of scientific research, such as physics, biology and astronomy data only by computer analysis may be fast enough. The world wide web (World Wide Web) is vast and growing, so retrieval of relevant information on the World Wide Web cannot rely on manual completion.
Machine learning, however, is not just a matter of databases, it is also a component of artificial intelligence. In order to be intelligent, a system in a changing environment must have the ability to learn. If the system can learn and adapt to these changes, then the designers of the system do not have to anticipate everything and provide solutions for them.
Machine learning uses instance data or past experience to train a computer to optimize some performance criteria. We have models that depend on certain parameters, and learning is the execution of a computer program that uses training data or previous experience to optimize the parameters of the model. Models can be predictive (predictive) for future forecasts or descriptive (descriptive),