How Do Machines Learn?
To keep things simple, just know that machines “learn” by finding patterns in data. Think of data as information you acquire from the world. The more data given to a machine, the “smarter” it gets.
But not all data is the same. Imagine you are a pirate and your life mission was to find buried treasure somewhere in the island. To find the treasure, you will need a substantial amount of information. Like data, this information can either lead you in the right direction or in the wrong direction. That's why it's important to be mindful of what kind of data you're giving your machine to learn. Some data may increase or decrease the uncertainty of your predictions.
Nevertheless, given a sufficient amount of data, the machine can make predictions. Machines can predict the future, as long as the future doesn't look very different from the past.
The machine actually "learns" using the old data to get information about what will most likely happen. If the old data looks a lot like the new data, then the things you can say about the old data will probably be relevant to the new data. It is like looking back to look ahead.
Read More: A Comprehensive Guide To Find A Right Data Science Job
types of machine learning
There are three main categories of machine learning:
Supervised learning: The machine learns from the labeled data.
Unsupervised learning: The machine learns from unlabeled data. Meaning, there is no "correct" answer given for the machine to learn, but the machine must expect patterns from the data to come up with the answer.
Reinforcement learning: The machine learns through a reward-based system.
Supervised machine learning
Supervised learning is the most common and studied type because it is easier to train a machine to learn with labeled data than with unlabeled data. Depending on what you want to predict, supervised learning can be used to solve two types of problems: regression or classification.
Regression problem: If you want to predict continuous values, such as trying to estimate how many hours a patient will stay in this hospital, you would use regression. This type of problem does not have a specific price constraint as it can be any number of days, hours or minutes.
Classification problem: If you are interested in a problem like: "Am I ugly?" So this is a classification problem because you are trying to classify the answer into two specific categories: yes or no (in this case the answer is yes to the question above).
Read More : Machine Learning As A Service Is Redefining The Businesses In New A Light!
Unsupervised Machine Learning
Since there is no labeled data for machines to learn from, the goal of unsupervised machine learning is to detect and group patterns in the data. Unsupervised learning is machines trying to learn "on their own" without help. Imagine someone throwing you a pile of data and saying "Here you go boy, find some patterns and group them for me. Thanks and have fun."
Depending on what you want to group together, unsupervised learning can group data together: clustering or association.
Clustering problem: Unsupervised learning tries to solve this problem by looking for similarities in the data. If there is a common cluster or cluster, the algorithm will classify them as a certain. An example of this might be an attempt to group customers based on past purchase behavior.
Association Problem: Inductive learning tries to solve this problem by trying to understand the rules and meanings behind different groups. An example of this can be seen when a store implemented this type of algorithm in its system. It turned out that there was a strong correlation between customers buying beer and diapers. He concluded from this statement that men who went out to buy diapers for their babies also tended to buy beer.
Conslusion
Reinforcement learning: The machine learns through a reward-based system.
0 Comments