Question 1

I have not done any machine learning, is this series good for me?

Accepted Answer

Although we do not require any specific ML concepts, we believe that the focus on deep learning is more useful to participants who are already familiar with some classical machine learning techniques as provided by the Python scikit-learn library.

Question 2

My data are in the form of multidimensional arrays with spatial patterns. Would this course apply?

Accepted Answer

Probably! In this series, by “image data” we mean any datasets that are at least two-dimensional, can have multiple channels, and have some visual patterns. It also includes other datasets which can be converted to this format: for example, audio signals can be converted to “spectrogram images”.

Question 3

I need help implementing a specific model from a recent paper. Will this series be useful for me?

Accepted Answer

Possibly not, this course is aimed as an introduction. We offer office hours with our data scientists which may be more helpful to you, see details here: https://escience.washington.edu/using-data-science/office-hours/

Question 4

I want to learn PyTorch.

Accepted Answer

We will use Keras with the tensorflow backend for the examples. Keras also can serve as a high level API for PyTorch. We will focus on conceptual topics that are transferrable across libraries, but if your primary  goal is to learn PyTorch syntax explicitly, you should find alternative tutorials.

Question 5

I am an R user, and I want to learn Python and Deep Learning! Will I keep up?

Accepted Answer

For this series we suggest basic familiarity with Python so that participants can interact with the tutorials. We will not teach Python syntax, so you will be expected to spend time on your own to catch up. The Keras library has an R interface, so what is learnt can be adopted into an R workflow.

Question 6

I am not a coder, but I want to learn how one can use deep learning in their research.

Accepted Answer

Although we will try to highlight the concepts behind the code, there will be a significant hands-on component, so this may not be ideal for you. You are welcome to join remotely to watch the presentations.

Question 7

Can I attend some tutorials in person and some remotely?

Accepted Answer

The tickets are separate and space is more limited in person, so please reserve a remote ticket if you are not able to attend all the tutorials in person.

Question 8

Can I attend only certain tutorials from this list?

Accepted Answer

For in-person tickets, we would like to give preference to people who will attend all tutorials. You are welcome to register to attend remotely for the tutorials of interest.

Pixel Proficiency: Practical Deep Learning for Images