cehr-bert
cehr-bert copied to clipboard
cumc-dbmi
CEHR-BERT: Incorporating temporal information from structured EHR data to improve prediction tasks
- Upload Tensorflow to 2.12 so we can use TensorFlow's version of MultiHeadAttention - Replace our own implementation of [MultiHeadAttention](https://github.com/cumc-dbmi/cehr-bert/blob/8862fd6a1f026839d301625fb30a1ba7e19a4672/models/custom_layers.py#L81) with the Tensorflow version of [MultiHeadAttention](https://www.tensorflow.org/api_docs/python/tf/keras/layers/MultiHeadAttention) - Toggle `mask` at...
Hi, Can you please add a simple notebook for a use-case of the models and going about training the network from scratch with sample Temporal data X, and prediction label...
I've been able to follow the tutorial in the README just fine, except for the last line. I've downloaded the Synthea data, converted it to the training data format, and...
The current normalization method for age is applied on the entire dataset before splitting up which might cause age information leakage for the future so it isn't aligned with the...
Adding an issue here to track https://github.com/Medical-Event-Data-Standard/MEDS-DEV/issues/134 Details: 1. The "direct" disk usage of CEHR-BERT on larger datasets (across all training stages) is pretty significant; on the denser dataset we're...
Hi, while using MEDS-DEV witht the HIRID dataset on a hypotension task I ran into this data error. The strange thing is that the pipeline runs for other tasks (e.g.,...
