Detailed method of machine learning

Machine Learning (ML) is a multi-disciplinary field that involves multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. Specializing in how computers simulate or realize human learning behavior to acquire new knowledge or skills, and reorganize existing knowledge structures to continuously improve their performance.

It is the core of artificial intelligence and the fundamental way to make computers intelligent. Its application covers all areas of artificial intelligence. It mainly uses induction, synthesis rather than deduction.

Comprehensively consider various historical factors such as historical origins, knowledge representation, reasoning strategies, similarities in outcome assessment, relative concentration of researchers, and areas of application. The machine learning method [1] is divided into the following six categories:

1) Empirical induc TIve learning

Empirical induction learning uses some data-intensive empirical methods (such as version space method, ID3 method, law discovery method) to induct examples. The examples and learning results are generally represented by symbols such as attributes, predicates, and relationships. It is equivalent to the inductive learning in the classification of learning strategies, but deducting the part of join learning, genetic algorithms, and reinforcement learning.

Machine learning

2) Analytical learning (analyTIc learning)

Analytical learning methods start with one or a few examples and use domain knowledge for analysis. Its main features are:

The reasoning strategy is mainly deduction rather than induction.

Use past problem solving experiences (examples) to guide new problem solving or generate search control rules that can use domain knowledge more effectively.

The goal of analytical learning is to improve the performance of the system rather than a new concept description. Analytical learning includes the application of interpretive learning, deductive learning, multi-level structural chunking, and macro manipulation learning techniques.

3) Analogy learning

It is equivalent to analogy learning based on learning strategy classification. The more compelling research in this type of learning is through analogy with specific examples experienced in the past, called case-based learning, or simply paradigmatic learning.

4) GeneTIc algorithm

Genetic algorithms simulate mutations, exchanges, and Darwin's natural selection of biological reproduction (the survival of the fittest in each ecological environment). It encodes the possible solution of the problem into a vector called an individual. Each element of the vector is called a gene and evaluates each individual in the group (set of individuals) using the objective function (corresponding to natural selection criteria). According to the evaluation value (fitness), individuals are subjected to genetic manipulations such as selection, exchange, and mutation to obtain new populations. Genetic algorithms are suitable for very complex and difficult environments. For example, with a large amount of noise and irrelevant data, things constantly updated, the problem objectives can not be clearly and precisely defined, and the value of current behavior can be determined through a long execution process. Like neural networks, the research of genetic algorithms has developed into an independent branch of artificial intelligence. Its representative figure is JH Holland.

5) Join learning

A typical join model is implemented as an artificial neural network consisting of some simple computational units called neurons and weighted joins between the units.

6) Reinforcement learning

Enhanced learning is characterized by the trial and error interaction with the environment to determine and optimize the choice of actions to achieve so-called sequence decision tasks. In this task, the learning mechanism causes the system state to change by selecting and executing actions, and it is possible to obtain some kind of enhanced signal (immediate return), thereby achieving interaction with the environment. Enhancing the signal is a scalar reward and punishment for system behavior. The goal of system learning is to find a suitable action selection strategy, that is, which action method to choose in any given state, so that the resulting action sequence can obtain some optimal result (such as the largest accumulated immediate return).

In the comprehensive classification, empirical induction learning, genetic algorithm, joint learning and reinforcement learning all belong to inductive learning, in which the empirical induction learning adopts symbolic representation, while the genetic algorithm, connection learning and reinforcement learning adopt sub-symbol representation; Deduction learning.

In fact, the analogy strategy can be seen as a synthesis of induction and deduction strategies. Therefore, the most basic learning strategies can only be summarized and interpreted.

From the point of view of learning content, learning using inductive strategies is based on the induction of input. The learned knowledge obviously exceeds the scope of the original system knowledge base. The learned results change the system's knowledge deduction and closure. The type of learning can also be called knowledge-level learning; while the learning using deductive strategies can improve the efficiency of the system, it can still be implied by the knowledge base of the original system, that is, the knowledge learned cannot be changed. The system deduction and closure, so this type of learning is also known as symbol-level learning.

Replacement Bulb Lamp With Housing

Replacement Projector Lamp with housing has one more case than the bare bulb lamp, which can be directly put into the projector and more convenient than the installation of the bare bulblamp. The projection effect is the same as the compatible projector lamp, with stable light source and brightness, which can support daily life use

Replacement Bulb Lamp With Housing,Bulb Lamp,Dynamic Projector Lamp,Led Projector Lamp

Shenzhen Happybate Trading Co.,LTD , https://www.happybateprojector.com

Posted on