Introduction to Machine Learning with Python

Python is considered one of the most simple and efficient programming languages. Its object-oriented programming approach and elegant syntax make it a powerful programming language. Python is an interpreted language. Its dynamic typing and high level data structures make it an ideal language for application development in various areas and on multiple platforms. Today, Python is widely used in the areas of machine learning and data science. The following chapter discusses Python, the utility of Python in machine learning and data science, ecosystem of Python in machine learning and various libraries in Python required for machine learning.

Machine Learning is referred to as the subset of Artificial Intelligence. It involves a machine being learned without programming it explicitly. In Machine Learning, machines try to improve their performance with the help of past experiences or by using training examples. The following chapter discusses Machine Learning, Selection of training set, Selection of target function, Selection of Function Approximation Algorithm, Perspectives and issues involved while building a machine learning system,In-sample and Out-of-sample error and Applications of Machine Learning.

Supervised learning is a machine learning task of mapping the input to the output on the basis of labeled input-output example pairs. Supervised learning may be of two types: classification and regression. In this chapter, we will discuss linear regression in one variable, linear regression in multiple variables, gradient descent, and polynomial regression.

Support Vector Machine (SVM) may be defined as a machine learning algorithm that can be used for regression and classification. It is generally used for classification purposes. In this chapter, we will discuss Margin and Large Margin Methods and Kernel Methods. 

Neural networks are a way to mimic the working of a human brain. Decision trees refer to the decision support structure that uses a tree to make decisions and draw all possible consequences. Decision trees are a way to display conditional control statements. In the following chapter, we will discuss decision trees, regression trees, stopping criterion and pruning loss functions in a decision tree, categorical attributes, multiway splits and missing values in decision trees, and instability in decision trees. 

A Support Vector Machine is used for the purpose of classification and regression. In the previous chapter, we discussed Support Vector Machine, margin and large margin methods, and kernel methods. A neural network refers to a parallel computing device that attempts to mimic the model of the brain. In the following chapter, we will discuss the early models of Neural Networks as well as the Perceptron learning model, back propagation, and Stochastic Gradient Descent.

Supervised learning involves training using well “labelled” training data, and on the basis of the training data, machines are able to predict the output. The labelled data means that the correct output is attached along with the corresponding input data. In supervised learning, as the name suggests, the training data acts as the supervisor and provides training to the machine to predict the correct output. This chapter discusses Statistical Decision Theory, Gaussian & Normal Distribution, Conditionally Independent Binary Components, Learning Beliefs Network and Nearest-Neighbour Methods.

Unsupervised learning is a complex processing task involving the identification of patterns in datasets having data points that are neither labeled nor classified. Unsupervised learning is a kind of machine learning algorithm that can be used to draw useful conclusions without the presence of labeled responses in the input data. In the following chapter, we will discuss Clustering (K-means Clustering, Hierarchical Clustering), and Principal Component Analysis.

Unsupervised learning is a kind of machine learning algorithm that can be used to draw useful conclusions without the presence of labeled responses in the input data. In the previous chapter, we discussed clustering (k-means clustering, hierarchical clustering) and Principal Component Analysis. In this chapter, we will discuss training versus testing, bounding the testing error, and the VC dimension.

In machine learning and AI, future predictions are based on past observations, and bias is based on prior information. Harmful biases occur because of human biases which are learned by an algorithm from the training data. In the previous chapter, we discussed training versus testing, bounding the testing error, and VC dimension. In this chapter, we will discuss bias and fairness.