Machine Reasoning

- Overview

Machine reasoning is a type of artificial intelligence (AI) that uses logic and rules to make inferences. It's a subset of AI that allows machines to act with agility and relevance based on learned data. 

Machine reasoning uses concepts and ideas coded as symbols, and then draws logical conclusions to try to resemble common sense. It can be seen as an attempt to implement abstract thinking as a computational system. 

Machine reasoning can:

• Lead a machine through the steps of a complex process.
• Adapt to real-time change.
• Execute functions.
• Enable machines to act with agility and relevancy according to learned data.
• Enable a computer to work through complex processes that would normally require a human.
• Deliver explainable recommendations.

Machine reasoning research aims to build inter-predictable AI systems that can solve problems or draw conclusions from what they are told and already know.

- Main Types of Reasoning in AI

Reasoning is a key component of machine learning (ML). It's the process of drawing logical conclusions from given information. In AI, reasoning is the ability of a computer to make deductions based on data and knowledge. 

Machine reasoning can be seen as an attempt to implement abstract thinking as a computational system. It applies human-like common sense to analyze and translate vast knowledge and learned network data into clear explainable insights. 

For example, a machine reasoning system can be fed an email that is potentially phishing, and draw its conclusions about whether the author is in fact suspicious.

Here are some types of reasoning used in AI: Deductive reasoning, Inductive reasoning, Abductive reasoning, Common sense reasoning, Monotonic reasoning, Non-monotonic reasoning.

- Machine Reasoning vs Machine Learning

Machine learning and machine reasoning are both part of AI and are important for decision-making and prediction modeling. However, they differ in how they process information and make decisions: 

• Machine learning: Uses large amounts of data to learn complex functions and make predictions. Machine learning systems are more formulaic and objective, and are good at reaching conclusions based on objective information. For example, a machine learning system might learn to classify images by analyzing their colors and the number of items in them.
• Machine reasoning: Uses logical techniques to generate conclusions from available knowledge. Machine reasoning systems are more open-ended and intuitive, and are good at reaching conclusions based on subjective and qualitative information. For example, a machine reasoning system might use a set of if-then rules to determine if a patient has a particular disease.

Machine learning and machine reasoning can be used together. For example, a machine reasoning system could identify suspicious purchases and then call on a machine learning system to create statistical models to help determine the rate of fraud.