Types of learning in machine learning explained

-

 Learning in the domain of machine learning is predominantly seen as in 4 buckets or types. Yes, there are 4 types of learning and in this article I will go through each one in a brief manner. For starters, the types of learning are:

  1. Supervised Learning
  2. Unsupervised Learning
  3. Semi supervised Learning
  4. Reinforcement Learning

Supervised Learning

Explanation

Supervised learning basically deals with labelled data. For example, let us say you have a flower species dataset in which you have flower characteristics (or features) like length of petal, length of sepal, width of petal and width of sepal and then you have an associated label (or target) which is the species of the flower. (This dataset can be found here.) 
The main point here is that the data is in the form of (X,y) where X is the set of input features (petal length, sepal length, petal width and sepal width) and y is the target which is flower species in this case.  In supervised learning, the target is known for the training data and you need to build a model using this training data which can predict the target (y) for unknown data where only the input features (X) are available. 
Furthermore, the target variable y can either be discrete or continuous. "Discrete" means discrete classes as in flower species or types of animals (dog, cat, wolf, etc.,). "Continuous" means that the target is a numerically ordered value like the price of a house. In supervised learning, if the target is discrete, it is called "Classification" and if the target is "continuous", it is called "Regression".

Why is supervised learning called so?

Supervised learning is called "supervised" learning because the whole training process happens under the supervision of the target variable or label. What does this mean? If you are aware of the model training process, it means that the model will predict the target value and then the label will supervise it to see if it is the correct label by comparing it to itself. If it is not, then this feedback will go to the training process for the next iteration and doing so in successive iterations, you will get a model with high prediction accuracy (which is why you have a no of iterations hyperparameter in ML library functions). You can think of it as having the solution key (label) for a question set with which you can compare your answers (predicted labels) after solving and then improve upon them in the training process.

Supervised learning
Supervised learning

Summary:

  • Supervised learning deals with labelled data.
  • Labelled data is in the form of (X,y) where X denotes input features and y denotes target variable. 
  • Aim is to train a model on this labelled data (X,y) so that it can predict unknown target y for some data X
  • Furthermore, supervised learning consists of 2 tasks: classification and regression
    • When the target y is discrete, it is called classification task.
    • When the target y is continuous, it is called regression task.

Unsupervised Learning

Explanation

Contrary to supervised learning, unsupervised learning deals with unlabelled data i.e., the data without any label or target. Unlabelled data means just X (set of input features) and no y (no target). Unsupervised learning is basically used to find patterns and form clusters in the available unlabelled dataset.

Why is Unsupervised learning called so?

Unsupervised learning is called "unsupervised" learning because unlike supervised learning there is no target feature which can supervise the correctness of the model process. You can think of it as giving an exam where the solution key to the questions is not available with you (so no supervision).

Unsupervised learning
Unsupervised learning


Summary

  • Unsupervised learning deals with unlabelled data.
  • The aim is to find patterns and cluster the unlabelled data.
  • K means clustering is a popular algorithm in this field.

Semi-supervised Learning

Explanation

Semi-supervised learning, as the name suggests, is a mixture of supervised and unsupervised learning. Basically, when you have a lot of data, but only a portion of that data is labelled and the rest is unlabelled, you can use the labelled data to infer the labels for the unlabelled data based on distance or something like that.

Why is semi supervised learning called so?

Semi-supervised learning is called "semi-supervised" because only a small portion of the data is labelled. So there is supervision but only weak or "semi" because of semi-labelled data.

Summary

  • Semi supervised learning deals with partially labelled datasets.
  • It is a combination of supervised and unsupervised learning.
  • An example could be pseudo labelling.

Reinforcement Learning

Explanation

In reinforcement learning, there is an agent in an environment and the agent takes some action and the environment rewards or punishes the agent, also sending it to a new state, thereby reinforcing certain behaviours of the agent which maximize its long term rewards (pretty similar to how humans act and learn when placed in a new environment).

Why is reinforcement learning called so?

The term "reinforcement" comes from behavioural psychology which basically means consequence of an action that strengthens certain behaviours of the organism (agent in this case). In reinforcement learning, we try to reinforce actions so as to maximize the long term rewards of the agent using some reward/punishment mechanism and that is why it is called so (similar to how your parents might reward you if you get good grades, and might punish you if you fail in hopes of reinforcing good habits in you).

Reinforcement learning
Reinforcement learning

Summary

  • Involves an agent acting in an environment which receives rewards or punishments according to its actions and these rewards/punishments shape its further trajectory of operation.
  • Similar to how most schools operate in the present world where you get rewarded for good grades and punished for bad grades so as to direct your behaviour towards good grades.

Popular posts from this blog

ML technical definition : Learning in AI defined

-
Image
Now that you have dived deep into the waters of machine learning, you might be interested in knowing how to define it. But before you can define machine learning, you need to learn what "learning" is. So, what is "learning" in Machine Learning? Learning in ML and AI  How do human beings learn? They improve with their experiences. Right!. This is exactly what learning is. But how would you define it, then? Learning can be defined as the ability to improve with experience. In terms of AI and ML, it means making computer programs and systems which automatically improve with experience like human beings do. Experience can be thought of as "data". In ML, we try to build models which learn from data to perform some task like decision-making, prediction, etc. But this seems like an informal definition of ML. Right! How would you define it formally? Definition of ML The present standard definition of machine learning is given by Tom Mitchell, as follows:  A comput...
Read more

Batch Gradient Descent vs Stochastic Gradient Descent (SGD) vs Mini-Batch Gradient Descent

-
Image
 In this article, we will explore the basic difference between Batch Gradient Descent, Stochastic Gradient Descent (SGD) and Mini-Batch Gradient Descent.  But first let us understand what an epoch means. Meaning of the word "epoch" In Deep Learning, an epoch is one iteration over the entire sample space (training set). In other words, everytime you go over each sample of the training set, it is considered one epoch. Basic Deep Learning training process Initialize model parameters: weights and biases with some random value. Go over objects of training set. Adjust your model parameters : weights and biases w.r.t some cost/loss fuction. Repeat from step 2 till desired threshold is reached. Batch Gradient Descent or simple Gradient Descent In Batch Gradient Descent, in each epoch we go throught entire training set and then adjust parameters. In every epoch, parameters are adjusted once which makes it unsuitable for large datasets. Stochastic Gradient Descent (SGD) In Stochast...
Read more

What exactly is Artificial Intelligence in simple words ? Explain AI in simple terms.

-
Image
  AI explained in simple language: Remember, when you were a child and you saw some animal with paws. You did not know what it was and the adults told you it was a dog. Then, you saw some other animal and you called it a "dog" but the adults corrected you as it was a "cat". From then on, you started to differentiate between dogs and cats. This is what we call human intelligence i.e., the ability to learn over time from your environment.  Now, what if I told you a computer could also do what you did. Just like adults taught you, you could  teach a computer to differentiate between a dog and a cat. Would you call the computer "intelligent"? This is precisely what Artificial Intelligence (in short AI) is. It is giving computers human-like intelligence so that they could do the tasks traditionally reserved for humans. But why do we need AI? Because they are computers and can perform much more than a human could without getting tired. In fact, you are already u...
Read more