When I started to dig into what Machine Learning education is available I was pleasantly surprised to find that there is a nice selection to choose from in the MOOCs (Massive Open Online Courses) area. There might be more but I've looked at the following from the big three. From my novice point-of-view none of them look easy. And none are Java based using python.
Udacity ML courses are in the Data Science category and not a specific AI section. Udacity offers a ML Nanodegree (i.e., not free but with help and special services) and for an additional cost a "Nanodegree Plus" which guarantees employment (wow!). The courses in partnership with Google, University of Washington, and Georgia Tech. There are a range of individual courses (mostly free) as well and all seem interesting. Topics in the Nanodegree range from predicting likely survivors of the Titanic disaster to training an autonomous car.
Coursera. On offer are six ML courses which cost $79. The course work looks interesting and is affiliated with the University of Washington. The courses start with predictive systems and ends with deep learning. They also offer a ML focused on neural networks from the University of Toronto. They also mention Stanford as a contributor.
edX. edX offers a number of ML and AI courses from MIT, UC Berkeley and Microsoft but they are not organized in sequence or theme or than their bucket of CS courses. They do have a number of courses that seem to bridge theory with current technology, such as Developing Intelligent Apps.