2
$\begingroup$

There are several good books about machine learning and there are many mathematical explanations for several classification models (in most books personally read commonly explain Naive Bayes, Logistic Regression and SVM).

Besides the Naive Bayes, Logistic Regression, SVM and several decision trees for which I understand both mathematically and graphically, by observing data instances as scatterplot, there are other existing models for which (so far) haven't found graphical explanation.

When dealing with classification a good starting point is to draw scatterplot for data instances and then decide which algorithm to use by observing how the points are distributed in space and which fuction would split observing data into classes the best, therefore I would be glad to find any good resource with the collection of gpraphically explained classification algorithms, perhaps on some real examples. In case anyone of you knows for some useful resource of this kind I would be very thankful for sharing the address/book or article title.

$\endgroup$
2
$\begingroup$

The book Pattern Classification by Duda and Hart now in a second edition with a third author does a lot of graphics like that. I particularly like the graphical 2 d displays that show why for bivariate normal class conditional densities, linear discriminants are best when covariances are the same and quadratic discrimination when covariances are not equal. Perhaps the best available book on the market about statistical learning is the book by Hastie, Friedman and Tibshirani which takes maximum advantage of color graphics throughout.

Links on amazon: 1. Hastie et al.. 2 Duda and Hart.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.