Machine Learning Algorithms for Data Scientists: An Overview

Machine Learning Algorithms for Data Scientists: An Overview

Machine Learning Algorithms for Data Scientists: An Overview

Preface

Machine Learning models are widely used in different fields such as Artificial Intelligence, Business, Clinical and Biological Sciences which includes self-driving cars, predictive models, disease prediction, genome sequencing, spam filtering, product recommendation, fraud detection and image recognition . It has gained importance due to its capabilities of handling large volume of data, prediction and classification accuracy and validation procedures.

Machine Learning models are built on the basis of statistical and mathematical algorithms. One important aspect of machine learning is it does not stick to standard algorithm throughout modeling process instead it learns from the data over a period of time and improves the accuracy of the model.   Classification and prediction tasks are carried out based on the characteristics, patterns and relationship of the features present in the data set. Machine learning model also forms the basis of Deep Learning models.

Machine Learning models involve supervised learning, unsupervised learning, semi supervised learning and reinforcement learning algorithms.

Data Scientists analyze, model and visualize data and provide actionable insights to the decision makers. Machine learning algorithms and tools help the data scientist to carry out these tasks with the help of software such R and Python.

This book provides an overview of Machine Learning models, algorithms and its application in different fields through the use of R Software. It also provides short introduction to R software for the benefit of users.

Author assumes the users have basic descriptive and inferential statistical knowledge which is essential for building Machine Learning models.

Data sets used in the books can be downloaded from the author’s website.

Vinaitheerthan Renganathan

Title: Machine Learning Algorithms for Data Scientists: An Overview

Author: Vinaitheerthan Renganathan

Publisher: Self Published

Edition: 1st Edition

Copyright: © 2021 Vinaitheerthan Renganathan 

Contents

Chapter 1:  Introduction

Chapter 2:  R Software

Chapter 3:  Data Preprocessing

Chapter 3:  Decision Tree Algorithm

Chapter 4:  Random Forest Algorithm

Chapter 5:  Support Vector Machine Algorithm

Chapter 6:  Naïve Bayes Algorithm

Chapter 7:  Artificial Neural Network Algorithm

Chapter 8:  Clustering Algorithm

Chapter 9:  Text mining algorithms

Chapter 10: Image processing and classification

Chapter 1:  Introduction

Machine learning term became a buzz word in recent times due to its capabilities, easy to implement characteristics and high computing power. Machine learning models uses data to learn and take decision related to classification, pattern recognition and prediction. Machine Models improve the task under consideration through learning from the data and evaluating the results based on the performance measures. 

Machine learning models are being applied in different fields to carry out the following tasks (not limited to)

Speech Recognition

Pattern Recognition

Computer Vision

Robotic Control

Disease progression

Fraud detection

Spam filtering

Network intrusion detection

Self-driving cars

Recommender systems

There are different types of Machine Learning algorithms are available. Supervised, Unsupervised and Reinforcement Learning Algorithms

Supervised learning algorithms involve training and test data sets. The algorithm is trained on the training data set which contains outcome variable which needs to be classified or predicted i.e. Class label in case of classification problem or predicted value in case of regression problem. In some cases test data set used in the ML algorithms will not contain the outcome variable i.e. class label or predicted values which will be used to test the predictive accuracy of the machine learning algorithms. One of the issues with the supervised learning algorithm is to get the class label for the training data set and it might not be available for a large dataset.

Unsupervised learning algorithm