|Data Science Tools|
Author: Christopher Greco
This book introduces some of the products and tools that can be used to carry out statistical analysis of datasets to discover more about underlying trends.
The book opens with a chapter on advice on choosing your data tools, followed by how to import data into the recommended tools. The advice is sensible, but on the whole fairly obvious - choose software that's easy to use, readily available and updated regularly. In practice, the author covers Excel and the OpenOffice spreadsheet, KNIME and R, and for each of the packages looks at how to use the software for data analysis tasks (or rather statistical tasks) such as confidence intervals, normal distribution, T-Tests, and linear regression. The data used is sourced from US Federal Government datasets.
The rest of the book works through a range of data analysis, or more accurately statistical, tests, showing how to perform the test in each of the four chosen tools. There's a chapter on 'descriptive statistics' - mean, mode, median, variance and standard deviation. The chapter also covers cumulative probability charts and T-tests.
Next comes a chapter on correlation, regression, and confidence intervals. The author takes you step by step through the analysis of tornado data for each task in the different tools.
A chapter on methods for specific tools is next, looking at power, multiple regression/correlation, Benford's law, filtering, and wordcloud analysis.
This is a useful book if you're wanting to learn about how to use classic statistical methods to find out underlying information in a large dataset. The thing to be careful of is not to think this is a book about big data analysis along the lines of a NoSQL database. We're not really talking modern data mining using techniques from artificial intelligence and clever query languages. Instead, this is mainstream statistical analysis. So long as that's what you want to learn, this is a good introduction.