Hands-on Machine Learning with Python

Here is the perfect comprehensive guide for readers with basic to intermediate level knowledge of machine learning and deep learning. It introduces tools such as NumPy for numerical processing, Pandas for panel data analysis, Matplotlib for visualization, Scikit-learn for machine learning, and Pytorch for deep learning with Python. It also serves as a long-term reference manual for the practitioners who will find solutions to commonly occurring scenarios.

The book is divided into three sections. The first section introduces you to number crunching and data analysis tools using Python with in-depth explanation on environment configuration, data loading, numerical processing, data analysis, and visualizations. The second section covers machine learning basics and Scikit-learn library. It also explains supervised learning, unsupervised learning, implementation, and classification of regression algorithms, and ensemble learning methods in an easy manner with theoretical and practical lessons. The third section explains complex neural network architectures with details on internal working and implementation of convolutional neural networks. The final chapter contains a detailed end-to-end solution with neural networks in Pytorch.

After completing Hands-on Machine Learning with Python, you will be able to implement machine learning and neural network solutions and extend them to your advantage.

WHAT YOU'LL LEARN

* Review data structures in NumPy and Pandas
* Demonstrate machine learning techniques and algorithm
* Understand supervised learning and unsupervised learning
* Examine convolutional neural networks and Recurrent neural networks
* Get acquainted with scikit-learn and PyTorch
* Predict sequences in recurrent neural networks and long short term memory



WHO THIS BOOK IS FOR

Data scientists, machine learning engineers, and software professionals with basic skills in Python programming.

Ashwin Pajankar holds a Master of Technology from IIIT Hyderabad, and has over 25 years of programming experience. He started his journey in programming and electronics with BASIC programming language and is now proficient in Assembly programming, C, C++, Java, Shell Scripting, and Python. Other technical experience includes single board computers such as Raspberry Pi and Banana Pro, and Arduino. He is currently a freelance online instructor teaching programming bootcamps to more than 60,000 students from tech companies and colleges. His Youtube channel has an audience of 10000 subscribers and he has published more than 15 books on programming and electronics with many international publications.

Aditya Joshi has worked in data science and machine learning engineering roles since the completion of his MS (By Research) from IIIT Hyderabad. He has conducted tutorials, workshops, invited lectures, and full courses for students and professionals who want to move to the field of data science. His past academic research publications include works on natural language processing, specifically fine grain sentiment analysis and code mixed text. He has been the organizing committee member and program committee member of academic conferences on data science and natural language processing.

Chapter 1: Getting Started with Python 3 and Jupyter Notebook

Chapter Goal: Introduce the reader to the basics of Python Programming language, philosophy, and installation. We will also learn how to install it on various platforms. This chapter also introduces the readers to Python programming with Jupyter Notebook. In the end, we will also have a brief overview of the constituent libraries of sciPy stack.

No of pages - 30

Sub -Topics

1. Introduction to the Python programming language

2. History of Python

3. Python enhancement proposals (PEPs)

4. Philosophy of Python

5. Real life applications of Python

6. Installing Python on various platforms (Windows and Debian Linux Flavors)

7. Python modes (Interactive and Script)8. Pip (pip installs python)

9. Introduction to the scientific Python ecosystem

10. Overview of Jupyter Notebook

11. Installation of Jupyter Notebook

12. Running code in Jupyter Notebook

Chapter 2: Getting Started with NumPyChapter Goal: Get started with NumPy Ndarrays and the basics of NumPy library. The chapter covers the instructions for installation and basic usage of NumPy.

No of pages: 10

Sub - Topics:

1. Introduction to NumPy

2. Install NumPy with pip3

3. Indexing and Slicing of ndarrays

4. Properties of ndarrays

5. Constants in NumPy

6. Datatypes in datatypes

Chapter 3 : Introduction to Data Visualization

Chapter goal – In this chapter, we will discuss the various ndarray creation routines available in NumPy. We will also get started with Visualizations with Matplotlib. We will learn how to visualize the various numerical ranges with Matplotlib.

No of pages: 15

Sub - Topics:

1. Ones and zeros

2. Matrices

3. Introduction to Matplotlib

4. Running Matplotlib programs in Jupyter Notebook and the script mode

5. Numerical ranges and visualizations

Chapter 4 : Introduction to Pandas

Chapter goal – Get started with Pandas data structures

No of pages: 10

Sub - Topics:

1. Install Pandas

2. What is Pandas

3. Introduction to series4. Introduction to dataframes

a) Plain Text File

b) CSV

c) Handling excel file

d) NumPy file format

e) NumPy CSV file reading

f) Matplotlib Cbook

g) Read CSV

h) Read Excel

i) Read JSON

j) Pickle

k) Pandas and web

l) Read SQL

m) Clipboard

Chapter 5: Introduction to Machine Learning with Scikit-Learn

Chapter goal – Get acquainted with machine learning basics and scikit-Learn library

No of pages: 10

1. What is machine learning, offline and online processes

2. Supervised/unsupervised methods

3. Overview of scikit learn library, APIs

4. Dataset loading, generated datasets

Chapter 6: Preparing Data for Machine Learning

Chapter Goal: Clean, vectorize and transform dataNo of Pages: 15

1. Type of data variables

2. Vectorization

3. Normalization

4. Processing text and images

Chapter 7: Supervised Learning Methods - 1

Chapter Goal: Learn and implement classification and regression algorithms

No of Pages: 30

1. Regression and classification, multiclass, multilabel classification

2. K-nearest neighbors

3. Linear regression, understanding parameters

4. Logistic regression

5. Decision trees

Chapter 8: Tuning Supervised Learners

Chapter Goal: Analyzing and improving the performance of supervised learning models

No of Pages: 20

1. Training methodology, evaluation methodology

2. Hyperparameter tuning

3. Regularization in linear regression

4. Regularization in logistic regression

5. Regularization in decision trees

6. Crossvalidation, K-fold cross validation

7. ROC Curve

Chapter 9: Supervised Learning Methods - 2

Chapter Goal: Learn more algorithmsNo of Pages: 15

1. Naive bayes

2. Support vector machines

3. Visualization of decision boundaries

Chapter 10: Ensemble Learning Methods

Chapter Goal: Learn the in-depth background of ensemble learning methods

No of Pages: 10

1. Bagging vs boosting

2. Random forest

3. Adaboost

4. Gradient boosting

Chapter 11: Unsupervised Learning Methods

Chapter Goal: Detailed theory and practically oriented introduction to dimensionality reduction and clustering algorithms

No of Pages: 20

1. Dimensionality reduction

2. Principle components analysis

3. Clustering

4. K-Means method

5. Density-based method

Chapter 12: Neural Networks and Pytorch Basics

Chapter Goal: Understand the basics of neural networks, deep learning, and Pytorch

No of Pages: 10

1. Introduction to Pytorch, tensors

2. Tensor operations

3. Exercises

Chapter 13: Feedforward Neural Networks

Chapter Goal: In-depth introduction to basic dense neural networks along with necessary mathematical background and implementation. (chapter might split into two while writing)

No of Pages: 20

1. Perceptron model

2. Neural network and activation functions

3. Multiclass classification

4. Cost functions and gradient descent

5. Backpropagation

6. Pytorch gradients

7. Linear regression with PyTorch

8. Basic dense network with PyTorch for regression

9. Basic dense network with Pytorch for classification

Chapter 14: Convolutional Neural Network

Chapter Goal: Explore details behind CNNs and implement two solutions for image classification

No of Pages: 20

1. Dense network for digits classification

2. Image filters and kernels

3. Convolutional layers

4. Pooling layers

5. CNN for digits classification

6. CNN for image classification

Chapter 15: Recurrent Neural Network

Chapter Goal: Understand sequence networks and implement them for forecasting values (or text classification)

No of Pages: 15

1. Introduction to recurrent neural networks

2. Vanishing gradient problem

3. LSTM

4. RNN batches, LSTM

5. Text classification Problem (or forecasting problem)

Chapter 16: Bringing It All Together

Chapter Goal: Discuss, conceptualize, design, and develop end to end

No of Pages: 201. Project 1

2. Project 2