Classification in Machine Learning : Discussions

By | March 15, 2021
Classification in Machine Learning

Before getting to know about classification, let’s discuss a little about machine learning itself. Machine Learning is a scientific field that studies algorithms and statistical models used by computer systems by relying on patterns.

Machine Learning also gives the system the ability to automatically learn and improve from the experience without being explicitly programmed. Machine learning focuses on developing computer programs that can access data and use it for self-study.

Then we enter into the discussion of what is classification? Classification can be defined as the process of predicting the class or category of an observed value or a given data point. The categorized output can take a form like “Black” or “White” or “spam” or “no-spam”. Mathematically, classification is the task of approaching the mapping function (f) of the input variable (X) to the output variable (Y). It belongs to supervised machine learning where the target is also provided along with the input dataset.


Classification in Machine Learning

Classification in Machine Learning

description of classification

An example of a classification problem could be spam detection in email. There are only two categories of output, “spam” and “no-spam”; then this is the binary type classification. To implement this classification, we first need to train classifiers. For this example, the “spam” and “no-spam” emails will be used as training data. After successfully training the classifier, it can be used to detect unknown emails.


As the name suggests, this kind of learner waits for the test data to appear after saving the training data. Classification is carried out only after obtaining test data. They spend less time training but more time predicting.

Example: K-nearest neighbor and case-based reasoning. · Eager Learners

In contrast to lazy learners, eager learners create classification models without waiting for the test data to appear after saving the training data. They spend more time training but less time predicting.

Example: Decision Tree, Naïve Bayes, and Artificial Neural Networks (ANN).


The following are several classification algorithms:

  • Logistic Regression


Some of the most important applications of the classification algorithm are as follows:

  • Speech Recognition

source :