Well, my recommendation is that you first try to learn the basics on SVM and Random Forrests. There are many tutorials available in the net. Videos on youtube are numerous. Try and think!
https://www.hackerearth.com/practice/...algorithms/tutorial-random-forest...r/tutorial/ ( Practical Tutorial on Random Forest and Parameter Tuning in R )
First you cannot choose classification technique randomly. It all depend on your type of data and applications. If one technique works good for one application then it may not necessary to work with another applications also. There are lots of good tutorials available for same.
You can use weka and LibSVM for the initial stage.
Do you wish to understand the methods? Than you need first to read a lot, try some use cases of classification, read again, return to the use cases, and so on.
Do you want onyl to learn how to performe classification? Then use any programm that has them implemented, and just play with one or two tutorials.
Keep in mind that SVM depends highly on fine tuneing and includes for free overfitting, but Random Forest has non of this. So if you are really new to this area of the world, try starting with Random Forest.