Prepare sc for py transform (#35)

* enable notebook on bigdl_0.3.0 and spark_2.2

* changes according to code changes

* remove changes on start_notebook.sh

* 0.3.0 notebook work

* Fix docs

* unsaved changes

* change README address for 0.3.0 download

* change README

* prepare sc for jenkins

* fix matplotlib import

* declare utf-8 coding

* more declares

* spark conf to avoid OOM

* change to directly use jupyter

* specify version 0.3.0 , delete spark install in Setup.md for pyspark will be installed with pip

* setup changes

* setup changes