Machine learning is a subset of artificial intelligence that includes techniques that enable computers to find problems in data and deliver artificial intelligence applications. Deep learning is a subset of machine learning that enables computers to solve more complex problems.
The operating environment of this tutorial: Windows 7 system, Dell G3 computer.
Artificial Intelligence (Artificial Intelligence), the English abbreviation is AI. It is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence.
Artificial intelligence is a branch of computer science that attempts to understand the nature of intelligence and produce a new intelligent machine that can respond in a manner similar to human intelligence. Research in this field includes speech recognition , image recognition, robots, natural language processing, intelligent search and expert systems, etc.
Artificial intelligence can simulate the information process of human consciousness and thinking. Artificial intelligence is not human intelligence, but it can think like humans and may even exceed human intelligence.
Data Mining, as the name suggests, is to “mining” hidden information from massive amounts of data. According to the textbook, the data here is “large and incomplete.” ", noisy, fuzzy, random practical application data", information refers to "information and knowledge that is implicit, regular, unknown to people in advance, but potentially useful and ultimately understandable". In a business environment, companies hope that the data stored in the database can "speak" and support decision-making. Therefore, data mining is more application-oriented.
Data mining is usually related to computer science and achieves the above goals through many methods such as statistics, online analytical processing, intelligence retrieval, machine learning, expert systems (relying on past rules of thumb) and pattern recognition.
Machine Learning refers to using certain algorithms to guide computers to use known data to derive appropriate models, and to use this model to provide insights into new situations. The process of judgment.
The idea of machine learning is not complicated. It is just a simulation of the learning process in human life. In this entire process, the most critical thing is data.
Any related research on learning algorithms trained through data belongs to machine learning, including many technologies that have been developed for many years, such as linear regression, K-means, prototype-based objective function aggregation class method), Decision Trees (Decision Trees, a graphical method using probability analysis), Random Forest (Random Forest, a graphical method using probability analysis), PCA (Principal Component Analysis, principal component analysis), SVM (Support Vector Machine, support vector machine) and ANN (Artificial Neural Networks, artificial neural network).
The concept of deep learning (Deep Learning) originates from the research of artificial neural networks. A multi-layer perceptron with multiple hidden layers is a deep learning structure. Deep learning discovers distributed feature representations of data by combining low-level features to form more abstract high-level representation attribute categories or features.
Deep learning is a new field in machine learning research. Its motivation is to build and simulate the neural network of the human brain for analysis and learning. It imitates the mechanism of the human brain to interpret data, such as images, sounds and text. .
Strictly speaking, artificial intelligence and machine learning are not directly related, but currently machine learning methods are widely used Just solve the problem of artificial intelligence. At present, machine learning is an implementation method of artificial intelligence, and it is also the most important implementation method.
Early machine learning actually belonged to statistics, not computer science; and the classic artificial intelligence before the 1990s had nothing to do with machine learning. So today's AI and ML have a lot of overlap, but there is no strict affiliation.
But if we only look at the computer department, ML belongs to AI. AI has become a very broad subject today.
Deep learning is a popular direction in machine learning. It is itself a derivative of the neural network algorithm and has achieved very good results in the classification and recognition of rich media such as images and speech.
So, if artificial intelligence and machine learning are regarded as two disciplines, the relationship between the three is as shown in the figure below:
If deep learning is regarded asAs a sub-discipline of artificial intelligence, the relationship between the three is shown in the following figure:
Machine learning is an important method of data mining, but machine learning is another discipline and is not subordinate to data mining. The two complement each other.
Source: http://m.elecfans.com/article/691751.html
The machine learning process uses the following steps Definition:
1. Identify relevant data sets and prepare them for analysis.
2. Select the type of algorithm to use.
3. Build an analytical model based on the algorithm used.
4. Carry out model training based on the test data set, and modify the model as needed.
5. Run the model to generate test scores.
The difference between machine learning and deep learning
1. Data volume:
Machine Learning can adapt to various data volumes, especially scenarios with small data volumes. On the other hand, if the amount of data increases rapidly, the effect of deep learning will be more prominent. The figure below shows the performance levels of machine learning and deep learning under different amounts of data.
2. Hardware dependency:
Contrary to traditional machine learning algorithms, deep learning algorithms are highly dependent on high-end equipment. Deep learning algorithms need to perform a large number of matrix multiplication operations and therefore require sufficient hardware resources to support them.
3. Feature Engineering:
Feature engineering is the process of putting domain-specific knowledge into specified features, aiming to reduce the level of data complexity and generate data that can be used for learning algorithms mode.
Example: The traditional machine learning model focuses on finding pixels and other attributes needed in feature engineering. Deep learning algorithms focus on other high-level features of the data, thus reducing the actual workload of the feature extractor for each new problem.
4. Problem-solving approach
Traditional machine learning algorithms follow standard procedures to solve problems. It breaks the problem into parts, solves them separately, and then combines the results to get the desired answer. Deep learning solves problems in a centralized manner without splitting the problem.
5. Execution time
Execution time refers to the amount of time required to train the algorithm. Deep learning takes a lot of time to train because it contains more parameters, so the time investment in training is also more significant. Relatively speaking, the execution time of machine learning algorithms is relatively short.
6. Interpretability
Interpretability is one of the main differences between machine learning and deep learning algorithms - deep learning algorithms are often not interpretable . Because of this, the industry will always think twice before using deep learning.
Practical applications of machine learning and deep learning:
Future development trends:
If you want to read more related articles, please visitPHP Chinese website! !
The above is the detailed content of What is the relationship between artificial intelligence, machine learning, and deep learning?. For more information, please follow other related articles on the PHP Chinese website!