Data Science with Python: Combine Python with machine learning principles to discover hidden patterns in raw data

by Rohan Chopra, Aaron England, Mohamed Noordeen Alaudeen

Leverage the power of the Python data science libraries and advanced machine learning techniques to analyse large unstructured datasets and predict the occurrence of a particular future event. Data Science with Python begins by introducing you to data science and teaches you to install the packages you need to create a data science coding environment. You will learn three major techniques in machine unsupervised learning, supervised learning, and reinforcement learning. You will also explore basic classification and regression techniques, such as support vector machines, decision trees, and logistic regression. As you make your way through chapters, you will study the basic functions, data structures, and syntax of the Python language that are used to handle large datasets with ease. You will learn about NumPy and pandas libraries for matrix calculations and data manipulation, study how to use Matplotlib to create highly customizable visualizations, and apply the boosting algorithm XGBoost to make predictions. In the concluding chapters, you will explore convolutional neural networks (CNNs), deep learning algorithms used to predict what is in an image. You will also understand how to feed human sentences to a neural network, make the model process contextual information, and create human language processing systems to predict the outcome. By the end of this book, you will be able to understand and implement any new data science algorithm and have the confidence to experiment with tools or libraries other than those covered in the book. Data Science with Python is designed for data analysts, data scientists, database engineers, and business analysts who want to move towards using Python and machine learning techniques to analyze data and predict outcomes. Basic knowledge of Python and data analytics will prove beneficial to understand the various concepts explained through this book.

426 pages, paperback

Published July 19, 2019

Reviews

[Zhijing · Rating 5 out of 5]

I actually find it a very good handbook. Similar books for machine learning beginners: Deep Learning.

Motivation: Apart from theoretical statistics knowledge, and fancy neural network models, most of the data analysis people working on real-world applications have to deal with messy and limited data. This book links class knowledge of ML to actual coding practices, and focuses a lot on simple but useful scenarios. (I found it echoing with my previous experience when helping physicians to analyze hundreds or thousands of medical data.)

Example contents:
- How to use pandas library to store data and clean data
- Data Transformation: A secret in building all types of models is to always keep in mind the data format and overview, and then pipe them into the normal input-features-to-embedding-to-output framework. Common formats: numerical (discrete and continuous), categorical, or vectors.
- Metrics: accuracy, F1, confusion matrix, precision, recall
- Reduce dimensions: PCA, LDA, HCA, etc.

I keep getting amazed at how versatile pandas, sklearn, and scipy are :).

[Man]

I'm only on chapter 1 so I don't feel like I can star rate this book yet, but I can already say that I do not recommend this book.

Why?
1. There are a lot of mistakes in it. It seems like no one proof read this book.
2. The concepts are not formalized and not communicated very well. (These guys are not writers and it shows. Their editor didn't do them any favors either.)
3. Exercise instructions are misleading in some places (you will notice a difference between the book and the provided materials).
4. From the beginning, the Anaconda environment yaml does not work because...
5. The book's resources are out on GitHub, but have not been maintained since 2019. Some of the required package versions are not available anymore.
6. The last several commits to the material repository out on GitHub all have the commit message "Changes"...which is more of a "yikes 🚩" than anything else.



With that out of the way, I just want to give some advice to people that have this book, but still want to try and use it. That's really why I'm here.

1. Install python packages into your Anaconda environment as you go. Obvious ones to start are: pandas, numpy, jupyter, matplotlib and ipython. I've also encountered seaborn and scikit-learn. Don't sweat the newer versions, you will just have to be on the lookout for warnings and adjust code around new method names and deprecation.

2. Sometimes these guys do silly things like make a shallow copy instead of a deep copy of a dataframe and it causes problems even in the jupyter notebooks they wrote themselves. Some intermediate understanding of programming and beginner understanding of Python will help you catch and correct this mistakes for your own hands-on practice. So make sure you're a strong programmer and brush up on Python features if you have to before you start this book.

3. Be ready to read documentation on your own. These authors don't provide much description of the libraries and methods they tell you to use. It's pretty much just: "Do this. Do this. Then do that." Which gives you the illusion that you've learned something. You will learn "how to," but not "why?" Instead, you will need to take subject detours to find out what is going on and how methods work, what arguments they take, and what they return.

Good luck. You can probably get set up and muddle through this book if this isn't your first time dealing with less than perfect manuals and documentation, but if you are a total beginner, there is probably a better textbook out there.