Human learning patterns are different from one person to another. Having a learning process embedded in people’s minds can be a challenge to change.
However, with Machine Learning (ML), the learning method can easily be modified simply by choosing a different algorithm. As for Machine Learning (ML), having well-defined processes at our disposal allows us to understand and assess the accuracy of learning. Evaluating human learning is generally done via exams which are not considered as a measure of intelligence. Let’s look more in detail at the difference between the human learning and machine learning processes.
What does each one of them mean?
And what are the differences and similarities between the two?
What is Human learning?
Human learning is all about observing things, recognizing a pattern, elaborating a theory or model which explains that pattern, and then putting that theory to the test and checking whether it matches most or all of the observations.
Learning is, basically, a model that represents a pattern within a collection of observations. Without a feasible model, there is no learning.
Consider any mathematical formula or physics equation or biological theory or economic theorem or chemical equation, they all describe a scheme of the physical or natural world. We could take Newton’s Laws of Motion as an example.
He studied the motion of physical bodies and interpreted their movement simply by explaining the forces that act on them. Newton uncovered the « pattern » or « relationship » between the forces affecting a body and the movement that occurs in response to these forces and expressed it in the form of his laws.
The core concept here is this: No model or learning reflects « true » or « absolute » reality. All models or learning represent only an approximation of the observed reality.
Therefore, we have to upgrade the model or learning if we come up against new observations regarding the phenomena we are studying.
What is Machine learning?
Machine learning gives computers the power to perform tasks traditionally done only by people.
Everything ranging from driving a car to translating speech, machine learning brings about an enormous expansion of artificial intelligence capabilities, enabling software to bring meaning to the chaotic and erratic, unpredictable real world.
So what exactly is machine learning?Technically speaking, machine learning refers to the process of teaching a computer system how to accurately make predictions based on the data it is fed with.
The predictions can include whether a fruit in a photo is a banana or an apple, whether a person crossing the road in front of a moving car can be spotted, or if the use of the word « book » in a sentence means a paperback book or a hotel reservation, if an email is spam, or to identify speech with enough accuracy to create captions for a YouTube video.
Perhaps the main distinction with traditional computer software is the fact that a human developer did not write code instructing the system on how to distinguish between the banana and the apple.
However, a machine learning model was trained to accurately identify the fruits through the processing of a large amount of data, presumably in this case a large number of images branded to show a banana or an apple.
Generally, machine learning falls into two major categories: supervised learning and unsupervised learning.
Supervised learning:
In essence, this approach focuses on teaching machines by example.
As part of their training in supervised learning, systems are subjected to huge volumes of labeled data, like pictures of handwritten numbers annotated to show which digit they match.
With sufficient examples, a supervised learning system will eventually be able to recognize groups of pixels and shapes related to each number, leading to the recognition of handwritten numbers, to reliably identify the numbers 9 and 4 or 6 and 8.
Yet the training of these systems typically involves large amounts of labeled data, as some systems need to be exposed to millions of examples to achieve a task successfully.
Unsupervised training :
On the other hand, unsupervised learning processes use algorithms that identify patterns in the data, searching for similarities that break down the data into categories.
A good example is Airbnb, which groups houses available for rent by neighborhood, as well as Google News, which groups articles related to similar topics on a daily basis.
However, the algorithm is not meant to separate out any specific type of data, it merely looks for data patterns that can be clustered by their similarities, or anomalies varying from one another.
Where is Machine learning used?
The use of machine learning systems happens all around us and is a mainstay of modern internet.
Machine learning systems serve to recommend a product you want to buy next on Amazon or a video you want to watch on Netflix.
With each Google search, several machine learning systems work together, ranging from understanding the language in which you’re searching to tailoring your results so that « bass » fishing enthusiasts are not swamped with guitar results. Likewise, Gmail’s spam and phishing recognition systems use auto-learning models to keep spam out of your inbox.
Among the most visible manifestations of the power of machine learning are virtual assistants, including Amazon’s Alexa, Apple’s Siri, Microsoft Cortana, and Google Assistant.
All of them strongly depend on machine learning in order to sustain their speech recognition as well as their skills to understand natural language, with an immense need for a corpus to answer questions.
In addition to these highly noticeable manifestations of machine learning, systems are starting to be used in almost every industry. Examples of such uses include : facial recognition for surveillance in countries such as China; computer vision for driverless cars, drones and delivery robots, speech and language recognition and synthesis for chatbots and service robots; providing assistance to radiologists to detect tumors with X-rays, Predictive maintenance of infrastructure through data analysis of IoT sensor data; guiding researchers to identify genetic sequences linked to diseases and identifying molecules that could lead to more effective drugs in healthcare ; computer vision support that makes the Amazon Go supermarket possible without a checkout; enabling reasonably accurate transcription and translation of speeches for business meetings; and the list is endless.
What is the difference between machine learning and human learning?
Both humans and machines make errors while using their intelligence during problem solving. In ML, systems store all the examples, if we take an overstocked model that has a poor generalization it will not work on any unprecedented examples.
Also, the education system in most Asian countries surpasses the students by providing coaching and classes on technical subjects which allows them to solve only exemplary problems. Such problems are solved in exams with no need for intelligence of any kind. Those students can solve problems they have already seen and only the problems they have seen in the past. They cannot correctly handle general problems with precision because their intelligence is not generalized. That is the main reason why skill levels of university graduates are insufficient. Simply put, the great majority of students become exaggerated models of learning.
Nowadays, machine learning is a booming part of the growing artificial intelligence research. This is due to the implementation of neural network software, imitating the human brain’s functions, as well as the availability of large and affordable computing hardware resources, which offers potential for solving problems that until now have depended on the power of the human brain.
Massive amounts of data (Big Data) consisting of medical or financial information, image libraries or information on customer behavior, and so on, are all processed by very complex types of algorithms to generate digital knowledge with no need for traditional programming.
When we compare machine learning to human learning theories, ideas become much easier to understand and less confusing. But obviously there remain some fundamental differences between the two, which is primarily the barrier that keeps artificial intelligence away from approaching general human intelligence.
References: