This post details the customizations I have made to the software I use. I thought it might be helpful to others.
I’m always curious about how people set up their computers and what software packages they have, so I thought I would share mine. So you know where I’m coming from, I use Windows, Mac, and Linux nearly every day. Part of my goal is to make transitioning between these systems as seamless as possible, but the exact setup varies by the operating system. For those of you on Windows, many of these can be installed with Ninite. I do a pretty good job of keeping this list up-to-date, so it should reflect my current recommendations for software.
Now that we’ve cleaned and prepared the data, let’s try classifying it using logistic regression. Logistic regression is a popular machine learning algorithm for classification due to its speed and accuracy relative to its simplicity.
Splice junctions are locations on sequences of DNA or RNA where superfluous sections are removed when proteins are created. After the splice, a section, known as the intron, is removed and the remaining sections, known as the exons, are joined together. Being able to identify these sequences of DNA is useful but time-consuming. This begs the question: Can spliced sections of DNA be determined with machine learning?
There are many ways to save and load models in TensorFlow and Keras. It’s good to have a range of options but sometimes with all of the flexibility it’s hard to know which one you actually need in the moment. This post demonstrates the different methods available and talks about the strengths of each.
In this notebook, we’re going to take the [dataset we prepared] and continue to iterate on the modeling. Last time we built a model using TensorFlow and Xception. This time, we’re going to iterate on that using FastAI.
