timeseriesAI is a library built on top of fastai/ Pytorch to help you apply Deep Learning to your time series/ sequential datasets, in particular Time Series Classification (TSC) and Time Series Regression (TSR) problems. The library contains 3 major components:
Notebooks: they are very practical, and show you how certain …
The UCR datasets are broadly used in TSC problems as s bechmark to measure performance. This notebook will allow you to test any of the available datasets, with the model of your choice and any training scheme. You can easily tweak any of them to try to beat a SOTA …
As you may know, Jeremy Howard claims in his excellent fastai course that data augmentation is perhaps the most important regularization technique when training a model for Computer Vision, second only to getting more data samples (which is often costly or just impossible).
During the last 2 years a number …
What I want to do is build a model for sequence-to-sequence type prediction on the Dow Jones Industrial Average (DJI). However, rather than treat this as a regression problem, I want to discretize the problem, so that what I am predicting is one of a set number of possible “types …