Over a million developers have joined DZone.
{{announcement.body}}
{{announcement.title}}

Demystifying AI and Machine Learning (Part 2)

DZone's Guide to

Demystifying AI and Machine Learning (Part 2)

Read this article in order to learn more about expert systems and neural networks in AI and machine learning.

· AI Zone ·
Free Resource

Insight for I&O leaders on deploying AIOps platforms to enhance performance monitoring today. Read the Guide.

This article is the continuation of the part 1 posted previously. In this article, I am explaining two key areas of focus in Artificial Intelligence. The aim of artificial intelligence is to make machines as intelligent as possible like human beings.

Expert Systems

Expert systems are knowledge based systems that rely on a knowledge base to solve a problem. A knowledge base can be represented in different forms such as rules, semantic networks, and decision trees. Expert systems consists of a knowledge base and an inference engine to infer or reason knowledge from the stored knowledge base. Expert systems are used in places where a human intelligence or human expert is needed to sovle a particular problem.

Knowledge Base

Rule based expert ssytems captures an expert's knowledge in a particular domain in the form of rules. These rules form the knowledge base, which are then evaluated against the facts by the inference engine to solve a particular problem. Example of a rule:

IF sky is clear

and sun is shining

THEN rain coat is not needed

Pros

  • Easy to capture an understanding of an expert's knowledge as rules are represented in natural language.

Cons

  • Experts vary in their opinions for the same topic, which makes it difficult to capture the domain knowledge.
  • Maintenance and updating of rules is a lengthy procedure.

Different types of expert systems exist like rule based expert systems, fuzzy expert systems, and frame based expert sysems.

Inference

Inferencing in expert systems happens through forward or backward chaining. Forward chaining is a data driven reasoning technique, which starts with knows data and proceeds forward with that rule. Backward chaining is a goal driven reasoning, which starts with a goal in mind and proceeds backwards to find the data that supports the goal.

Neural Networks

Artificial Neural Networks (ANN) are inspired from the natural neural network of human nervous system. The system is made to work exactly the way a human brain stores and processes knowledge. A neural network very similar to the human brain is composed of a set of neurons or nodes highly connected to each other. Information is stored, processed, and analysed in the neurons of a network. Each node or neuron can activate other neurons in the network. The link or connection between neurons is called weights. A network can contain n number of neurons or nodes, which can make the network really complex. A simple neural network consists of a single input and output layer.

The following are the different types of neural networks:

  • Feed forward neural network
  • Convolutional Neural Networks (CNN)
  • Recurrent neural networks
  • Long Short-Term Memory networks (LSTM)

Artificial neural networks are capable of learning through adjustment in their weights. It's this capability of netural networks that makes them suitable of machine learning. Different types of learning algorithms can be used with neural networks. The most prominent of this is the back-propogation training or learning algorithm.

TrueSight is an AIOps platform, powered by machine learning and analytics, that elevates IT operations to address multi-cloud complexity and the speed of digital transformation.

Topics:
ai ,neural network ,artificial neural network ,machine intelligence ,research ,expert systems ,machine learning ,artificial intelligence

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}