search

Leezekun / ViTST

[NeurIPS 2023] The official repo for the paper: "Time Series as Images: Vision Transformer for Irregularly Sampled Time Series"."
https://arxiv.org/abs/2303.12799
111 stars 8 forks source link

How to adapt framework for different type of data? #2

Open gabriead opened 1 year ago

gabriead commented 1 year ago

Hi team, great work first of all. I would like to use your framework for the domain of sports. I have a time series data of players in regards to different parameters. How would I use the data witht the current model? In what input format does the data have to be in? Is a dataframe sufficient that is grouped by player and his parameters? Can the model be trained on the data of an entire team and then predict a parameter of a given player? Thank's a lot in advance!

Leezekun commented 1 year ago

Hi,

Thank you for your patience and for showing interest in our work.

You can locate the code that we've developed to convert data into line graph images with matplotlib in dataset/{}/process_scripts/ConstructImage.py. I recommend checking out the code to understand the expected input format. Alternatively, if it suits you better, you can craft your own code to create images using matplotlib or other methods.

As for your specific task, you can definitely give it a try.

Warm regards, Zekun

ericleonardo commented 7 months ago

Hi,

Thank you for your patience and for showing interest in our work.

You can locate the code that we've developed to convert data into line graph images with matplotlib in dataset/{}/process_scripts/ConstructImage.py. I recommend checking out the code to understand the expected input format. Alternatively, if it suits you better, you can craft your own code to create images using matplotlib or other methods.

As for your specific task, you can definitely give it a try.

Warm regards, Zekun

Hi Zekun... What you think about financial market prediction (stock, forex, etc) considering they are regularly sampled? Is there any negative issue in using regularly sampled time series with ViTST? In case, the goal is to predict/classify if price will go Up or Down in next step.

Thank you and congratulations!