Explain the basics of scikit-learn library in Python?

Scikit-learn, commonly known as sklearn is a library in Python that is used for the purpose of implementing machine learning algorithms.

It is an open-source library hence it can be used free of cost. Powerful and robust, since it provides a wide variety of tools to perform statistical modelling. This includes classification, regression, clustering, dimensionality reduction, and much more with the help of a powerful, and stable interface in Python. This library is built on Numpy, SciPy and Matplotlib libraries.

It can be installed using the ‘pip’ command as shown below −

pip install scikit-learn

This library focuses on data modelling.

There are many models used in scikit-learn, and some of them have been summarized below.

Supervised Learning Algorithms

Supervised learning algorithm is taught to behave in a certain way. A certain desirable output is mapped to a given input thereby providing human supervision. This could be by labelling the features (variables present in the input dataset), by providing feedback to the data (whether the output was predicted correctly by the algorithm, and if not what the right prediction has to be) and so on.

Once the algorithm is completely trained on such input data, it can be generalized to work for similar kinds of data. It will gain the ability to predict results for never-before-seen inputs if the model that is trained has good performance metrics. It is an expensive learning algorithm since humans need to physically label the input dataset thereby adding to additional costs.

Sklearn helps implement Linear Regression Support Vector Machine, Decision Tree, and so on.

Unsupervised Learning

This is opposite to supervised learning, i.e. the input data set is not labelled, thereby indicating zero human supervision. The algorithm learns from such unlabelled data, extracts patterns, performs predictions, gives insights into the data and performs other operations on its own. Most of the times, real-world data is unstructured and unlabelled.

Sklearn helps implement clustering, factor analysis, principal component analysis, neural networks, and so on.


Similar data is grouped into a structure and any noise (outlier or unusual data) will fall outside this cluster which can later be eliminated or disregarded.

Cross Validation

It is a process in which the original dataset is divided into two parts- the ‘training dataset’ and the ‘testing dataset’. The need of a ‘validation dataset’ is eliminated when cross-validation is used. There are many variations of ‘cross-validation’ method. The most commonly used cross-validation method is ‘k’ fold cross-validation.

Dimensionality Reduction

Dimensionality reduction tells about the techniques that are used to reduce the number of features in a dataset. If the number of features are higher in a dataset, it is often difficult to model the algorithm. If the input dataset has too many variables, the performance of machine learning algorithms can degrade by a considerable amount.

Having a large number of dimensions in the feature space requires large amount of memory, and this means not all of the data can be aptly represented on the space (rows of data). This means, the performance of the machine learning algorithm will be affected, and this is also known as the ‘curse of dimensionality’. Hence it is suggested to reduce the number of input features in the dataset. Hence the name ‘dimensionality reduction’.

Updated on: 11-Dec-2020


Kickstart Your Career

Get certified by completing the course

Get Started