Jupyter kernel keeps dieing/disconnecting

It's always down to 2 things: out of memory, or TensorFlow explosion. Avoid posting +1's/bumps, please. The forum is supported by the community and responses can take 2-7 days. 

When you say "Tensorflow explosion" are you referring to gradient explosion? If I'm not seeing that in the loss calcs by epoch are you saying that I need a larger instance?

Hi Nathan,

Tensorflow explosion would be gradient explosion. Although the loss may give reasonable values, if the weights are on the edge of being too small or too large, if they are nudged, then the numerical instability can cause overflow/underflow.Here are some suggestions to address the out-of-memory issue:

1. If there is numerical instability, I'd suggesting clamping (setting a min floor) or clipping (setting a max ceiling)
2. Use more layers with less nodes in each - this can both reach the solution faster as well as reduce the number of hidden parameters. Feel free to also try just using less nodes
3. Apply standardization of the data to the Standard Normal distribution - this can help prevent overflow if overflow occurs
4. Apply dimensionality reduction methods such as PCA

Also, please keep in mind if the model causes memory issues, then there could be overfitting.

Best,
Shile Wen

Hi Shile, 

>>if the model causes memory issues, then there could be overfitting

Could you explain?

Hi Valery,

If the model needs so much resources that training causes memory issues, then the model could be overly complex, meaning it is affected heavily by slight noise. Given that the markets are very noisy, we'd prefer a biased model (see Bias-Variance Tradeoff) to capture general movements.

Best,
Shile Wen