S 14 along 27 In March, there is a spring sale in the publishing house Peter, when all paper books are discounted. 50% by coupon - Book
And although I don't like the translations of this publisher, some books they do well, for example, the book "Theoretical minimum on Big Data" from the guys with complex names Annalin Eun and Kenneth Su. The book is a fire for those who want to learn about basic methods of processing big data, but do not want to remember mathematics outside of high school:)
The authors first consider some basis:
- preparation of data (Data format, types and choice of variables, feature construction and incomplete data)
- choice of algorithm (learning without a teacher, learning with a teacher, learning with reinforcement)
- setting parameters (How to deal with overeducation and undereducation n) evaluation of results (metrics and validation) and further the methods in the format description of the method, example of use, method limitations and outputs.
The book discusses methods:
- clustering by the method of k-averages on the example of profiles of coinoperators
- the method of the main components on the example of the study of nutritional value associative rules on the example of search for buying patterns Analysis of social networks on the example of the relationship scheme:) Regression analysis on the example of estimating the price of a house The K-Nearest Neighbors Method for Wine Expertise
- the method of support vectors on the example of the detection of cardiovascular diseases The decision tree on the example of the rescue forecast from the Titanic Random forests on the example of predicting criminal activity Neural networks on the example of number recognition a/b tests and multi-armed bandits on the example of advertising spinning
In general, the book is very good for an initial acquaintance with all these methods. I recommend it.
#Software #BigData #ForBeginners