Calverley Consulting LLC Project Management

Artificial Intelligence and Machine Learning
By:  Melanie Calverley

Artificial Intelligence (AI)

Artificial Intelligence (AI) involves the science of making computers or programs that simulate the characteristics of human intelligence, thus making machines and software mimic the thinking processes of the human brain. Artificial Intelligence is a broad and complex science, using various methods to simulate human intelligence, such as machine learning. Behaviors such as problem-solving and learning, strongly signify human thinking patterns associated with human intelligence. With Artificial Intelligence, these behaviors are learning and achieved by analyzing datasets and successfully identifying patterns.

A solid example of Artificial Intelligence can be found in Siri or Alexa. These computers act as intelligent virtual assistants to humans in order to increase production. These assistants communicate with humans and use human speech and verbal capability, which makes them appear more humanoid.

Machine Learning (ML)

Machine Learning (ML) is a subset of Artificial Intelligence that studies computer algorithms in an effort for the program to learn from both small and large datasets through the examination and comparison of common patterns within the data. The Machine Learning model will analyze extensive datasets and patterns and eventually “learn” the patterns to identify and assign specific characteristics to a function.

Machine learning is based on the concept that machines will learn and then adapt through experience. For example, machine learning can be used to examine large and extensive datasets that would be difficult, if not impossible, for a human to decipher and utilize effectively. Machine learning has the capability to take the datasets and learn the data in a fraction of the time that it would take a human to analyze and process a portion of it. Artificial Intelligence, in contrast, has a function to appear “smart” and “intelligent” with the likeness to the way that the human brain synthesizes data. This is achieved by the machine or program executing tasks in a “smart” manner, similar to human thinking and execution.

Deep Learning

Machine Learning consists of several of its own subsets, including deep learning, supervised learning, unsupervised learning, and reinforcement learning, or even hybrids of blended characteristics of two or more of these subsets. Deep learning is a broader spectrum of machine learning methods that are based on artificial neural networks that mimic the human brain and its thought processes. Machines and programs can solve convoluted problems using deep learning characteristics. What makes deep learning unique is that it requires an enormous amount of labeled data to be effective.

Automated driving has taken flight in recent years, and companies like Tesla, are among the frontrunners in the industry. Automotive scientists, engineers, and researchers use deep learning to enable vehicles to automatically detect roads, markings, pedestrians, and other objects. More recently, for example, Tesla implemented its deep learning strategies into its vehicles to detect stop signs and traffic lights in real-time. This type of feat is only possible through countless hours of video and millions of images utilized through high levels of computing via powerful graphics processing units (GPUs) and analyzation.

Supervised Learning

Supervised learning, another subset of machine learning, uses algorithms to identify relationships and dependencies between existing datasets in order to execute a predictive pattern based on those datasets and their dependencies. In other words, supervised learning simulates an algorithm that is trained and it is largely based on the learned patterns resulting from the previous datasets.

Supervised learning, due to its need for complete data to train the algorithm, can be found in devices such as automated search engines and smart applications. In the real estate industry, for example, companies have developed programs that predict real estate prices for properties and rentals. Using extensive and organized datasets that contain historical and actual values of properties in specific areas, as well as other factors such as proximity to certain transportation entities and entertainment venues, a program using supervised learning can predict the cost of monthly rent in a certain city. Supervised learning allows the algorithm to successfully complete such an exhaustive task that would take humans months to accomplish through data mining.

Unsupervised Learning

Unsupervised learning, another subset of machine learning, works from a dataset that is not necessarily organized, labeled, or structured in a particular manner. Unlike supervised learning, where datasets are labeled or categorized, unsupervised learning utilizes algorithms that lack structure.

An example of unsupervised learning is a program that detects a user’s buying patterns and then provides a suggestion of another product to purchase. The algorithm is not necessarily working with organized and compartmentalized data, and the algorithm has not been given explicit instructions on what action to take with the data; however, utilizing unsupervised learning, the algorithm’s neural network will seek some kind of structure and pattern within the data.

Reinforcement Learning

Reinforcement learning, another type of machine learning, will analyze data from a reward and risk standpoint. A machine or program using reinforcement learning will aim to find an optimal method to accomplish or complete a task or goal. The same method can be used to improve performance on a specific task or set of tasks.

Video games are some of the best examples of reinforcement learning, and they can mimic the similarities of reward based on a current environment that is constantly seen in the animal kingdom. In other words, based on the current situation, an action can result in an optimal result. Feedback offered to video game players enables the players to identify strategies in their future actions during the games. For example, a video game player may be more aware of a potential pitfall in a similar-looking location to the one that he or she saw previously prior to losing in the game. Getting to the next level may be easier once the player understands the consequences of a certain action taken at a previous level of the game.


As we continue to develop and learn about Artificial Intelligence and Machine Learning, the advances in science and its benefits will continue to be a part of everyday life. From self-driving and autonomous vehicles to facial recognition software, the advances continue to gain strength in popularity. Not only does Artificial Intelligence and machine learning provide convenience and security in our world, but it also attributes to developments and advances in defense and medical science as well.

© 10/30/2020

Original article by author, Melanie Calverley, first posted by Leopard Project Controls at:

This is the title of the web page
Copyright © 2017 - 2023 Melanie Calverley, Calverley Consulting, LLC. All Rights Reserved. This website uses cookies to enhance your browsing experience.
This is the title of the web page