Introduction
Classification is one of the most important and widely used tasks in machine learning. It is a supervised learning approach in which the computer program learns from the data given to it and then uses that learning to classify new data. Classification is an essential part of many machine learning applications including facial recognition, spam detection, medical diagnosis, etc. So in this article, we will be having a complete overview of Classification in Machine Learning, what is it? and why it is important?
What is supervised learning?
Before understanding classification, we must first understand the different types of machine learning methods. There are supervised and unsupervised learning methods. Supervised learning is when the algorithm learns from labeled data that includes both input and desired output. Classification is a supervised learning method that takes input data and assigns them to one of several possible categories or classes, making it one of the most common machine learning tasks.
In order for an algorithm to classify something, there need to be two parts: (1) the algorithm itself and (2) training data with labels. Training data consists of examples that have been manually labeled by us with their correct category. The labels should be clear enough so as not to confuse an individual looking at them; if they're too ambiguous then accuracy can suffer. To train an algorithm, you would give it example inputs, let it analyze them using its functions, and provide the desired output.
 |
| Supervised Learning |
What is classification?
Let's look at the basic meaning of the term Classification. Classification in its exact meaning is a method of sorting things into groups. The simplest example of this is putting books on a shelf. You put all the books about animals on one shelf, all the books about plants on another shelf, and so on. Classification also predicts which category or class an object belongs to based on known information about what objects are in which category or class.
It's like teaching a kid. You first show him/her a bunch of pictures of animals and tell him/her what is this. This is a cat, this is a dog, this is a bird, etc. You then give him/her a new picture and he/she should be able to tell you what it is. If he/she cannot then you show him/her more pictures until he/she gets it right. Classification in Machine Learning also works in a similar manner by giving the system some data from which it can learn. For example, if your goal is to classify faces into different categories such as Male and Female, you would provide your system with lots of pictures of male faces and female faces. Now if I showed your system a picture that had both male and female features, your system would say I don't know. Your machine learning algorithm would have classified that face as unclassified because there wasn't enough data available for it to make a prediction.
 |
| Classification of Animals |
Classification Models
Classification models are basically the algorithms that are used to predict if a particular observation belongs to a class or not. In other words, classification models help us classify observations into different classes. They come in various types such as Random Forest Classifiers, Naïve Bayes Classifiers, Support Vector Machines, K-Nearest Neighbors, etc. Depending on their type, they can be trained using various data sets.
Some classification models,
K-Nearest neighbors
K-Nearest Neighbors or simply KNN works by assuming the similarity of an object to another based on a number of other objects it is similar to. For example, if you are trying to predict whether a person will be a customer of your product, you could base that prediction on whether other people who have purchased your product have been customers. This is a very simple algorithm, but it can be extremely effective in situations where there are not many variables to take into account.
Random Forest Classifier
The Random Forest Classifiers are capable of handling multiple classes directly. It is also quite efficient in terms of memory usage, which makes it a good choice for problems where training sets are very large. This algorithm can be used for both regression and classification problems.
Naive Bayes Classifier
The algorithm uses Bayes’ theorem to predict probabilities for unseen data based on training data. The model can be easily understood by considering an example: Suppose we have a training set consisting of 2 classes - 'spam' and 'not spam'. We are given features like word frequencies, email addresses, etc., for both these classes. Based on our training set, we can predict what will be the probability of a new email being spam or not spam using a naive Bayes classifier algorithm.
Support Vector Machine
Logistic Regression
Logistic Regression could be one of the most popular and powerful algorithms in machine learning. It is used to predict a binary outcome (0 or 1) based on continuous input variables. This algorithm is often used in spam detection, handwriting recognition, face detection, and credit card fraud detection. The key thing about logistic regression is that it can learn non-linear relationships between input variables and an output variable. For example, you can use logistic regression to determine if a person will buy your product or not based on their age, income level, gender, etc
Check this article to know more about different algorithms and their applications
Classifier Evaluation Metrics
When talking about Classification we need to also discuss the different metrics to evaluate a classifier. Some of these include accuracy, precision, recall, and the F-score.
Precision:- Precision is defined as the number of true positives divided by the total number of predicted positives.
Precision = TP / (TP + FP) (where TP is True Positives and FP is False Positives)
Recall:- The recall is defined as the number of true positives divided by the total number of all possible positives.
Recall = TP / (TP + FN) (where FN is False Negative)
Accuracy: -The accuracy is defined as (TP+TN)/(TP+TN+FP+FN). The F-score combines both precision and recall with ((2*precision*recall)/(precision+recall)).
F-Score:- It is often used as a metric for evaluation because it takes into account both precision and recall at the same time. The F-score combines both precision and recall with ((2*precision*recall)/(precision+recall)).
Follow this article to know more about classification metrics including precision and recall
Conclusion
In this article, we have seen how classification can be used to categorize data into groups. Furthermore, we’ve seen the different types of classification models including KNN, Random Forest, Logistic Regression, etc. We hope you found this blog post informative and helpful. If you have any queries, please mention them in the comment section.
