|Enrol Now For Udacity Intro to Self-Driving Cars Nanodegree|
|Written by Nikos Vaggalis|
|Monday, 21 May 2018|
Self-Driving Cars;sounds seriously challenging and not for mere mortals. Could it be otherwise? Yes, and that's what Udacity is trying to change with this nanodegree, to render the topic of self driving cars accessible to just about anyone. Enrollment for the next session ends on May 22.
The big turn off when trying to get into Artificial Intelligence and its components of machine and deep learning is the fearsome math that is required, especially algebra calculus and matrix algebra. At Udacity the course designers are well aware of this struggle and have set as prerequisite for the course that students be at least familiar with high school level maths. The course will do its best to advance you beyond this level, albeit in a very smooth way.
Programming wise then, you apply that math to writing programs, first in Python, and at the later and more advanced stage in C++. It's not a course on learning Python or C++ per se, but nevertheless does so indirectly by walking you through a lot of code demonstrations and practical exercises. However, familiarity with the basic programming concepts is strongly recommended in order to be able to follow along unhindered.
Self-driving cars, and most recently self-driving flying cars are, of course, a topic that Udacity's founder Sebastian Thrun has been interested in for decades. Udacity has offered a course on the topic from the outset and indeed it was a topic included in the 2011 Introduction to Artificial Intelligence class. We have previously looked at the advanced level Self-Driving Car nanodegree and now we turn attention to the introductory level nanodegree that starts on May 22nd and can be completed in 4 months if you devote 10 hours per week to it. Being just one term in length it costs $800.
The preview of the nanodegree, on the topic of Localization, represents a tiny portion of the full program experience - but a crucial one since it deals with finding out a self-driving car's location in the world, a quintessential act in order for the car to navigate safely.
In getting to know Localization the preview goes through the topics of Probability, Bayes' Rule, Kalman Filters and Object-Oriented Programming, with all the maths and programming that go together with them.
To quickly outline the preview's curriculum:
Lesson 2 - Probability starts with a high-level overview and then specifically looks into Probability in Robotics. Also there's Conditional Probability, Estimating Based on Conditions, Simulating Probabilities and a showcase of Probability of Collision through Python code. All code is tested in the 'PlayGround', a cloud IDE based on Jupyter, so there's no need to download anything.
Lesson 3 - Bayes' Rule's theoretical part is on Robotics, Learning from Sensor Data, Using Sensor Data and so on while the programming part has exercises on Programming Probabilities, Programming Bayes' Rule, Array Iteration, 2D Arrays and the Robot World.
Lesson 4 - Kalman Filters and Object-Oriented Programming
Finally there is also a project, the Histogram Filter, which can be previewed, but have to enrol in order to actually tackle it.
It's certainly exciting that such a course aims to make an advanced topic such as this approachable. What's more exciting is that the knowledge obtained in the concepts, maths, programming and machine learning are universal with far more reaching applications than the narrow field of self driving cars. You can then forward into things like Robotics, Autonomous systems, Drones or even Flying cars!
Enrollment ends on May 22, 2018
or email your comment to: email@example.com
|Last Updated ( Monday, 21 May 2018 )|