Demystifying AI and Machine Learning (Part 1)
Get an introduction to AI (a machine’s ability to make decisions and perform tasks that simulate human intelligence) and ML (a facet of AI that focuses on algorithms).
Join the DZone community and get the full member experience.Join For Free
This is the first part of series of articles on AI and machine learning.
Artificial Intelligence (AI)
Intelligence is the ability to think humanly and act rationally. Humans have the intelligence to think and to make decisions by applying heuristics and common sense. Artificial intelligence is a field of study related to building intelligent agents so that agents can think like a human and act rationally. An agent is any machine that can do some intelligent task. The Turing Test, proposed by Alan Turing (1950), was designed to provide a satisfactory operational definition of intelligence. An agent/bot can pass the Turing Test if it has the following capabilities:
Understand and write natural language to interact with a human.
Knowledge representation (knows how to present knowledge to a user).
Knowledge inference (knows how to infer answers from stored knowledge to answer humans).
Machine learning to extrapolate patterns and adapt to new circumstances.
AI, in short, is about the study of the above disciplines and algorithms that help build an intelligent agent. The set of problems solved by AI are NP-complete, which cannot be achieved in polynomial time.
AI is a vast area of research with the following five important disciplines:
Natural language processing
Machine Learning (ML)
Machine learning is a subset of artificial intelligence that deals with the study of algorithms that can learn from data by learning from analogies. Learning can turn a human into a genius and allow them to adapt to new environments. In the same way, the learning capability of an agent/bot makes it robust enough to adapt to new environments. How fast and how deep an agent can learn is a big area of research that has been taking place over many years. The goal of any machine learning algorithm is to maximize its objective through the learning process so that it can handle unseen circumstances.
Two key learning methodologies to achieve machine learning are:
Supervised learning: An external teacher or labeled data help the machine learn.
Unsupervised learning: The machine learns without any labeled data or external teacher.
Opinions expressed by DZone contributors are their own.