Christopher Teubert

Hyperparameter optimization

1

**Requirement Text** Add hyper parameter optimization example or feature **Background Information** Hyperparameter optimization is an important step in building ml models. It will be part of the expected workflow **Suggested...

**Requirement Text** Update lstm to use auto regression **Background Information** See https://www.tensorflow.org/tutorials/structured_data/time_series Auto regression will use the output form one step in the input to the next, similar to how...

https://keras.io/api/utils/python_utils/#progbar-class

**Requirement Text** Rework normalization to be configurable **Suggested Solution** Normalization layer **DoD** - [ ] Implement feature - [ ] Add to example - [ ] Add tests

**Requirement Text** Add ability to initialize a BatteryCircuit model using an electrochem EOD model **Background Information** Battery circuit and electro chem eod represent the same system in different ways. There...

**Requirement Text** Implement k-fold validation **Background Information** See page 135 of deep learning with python **DoD** - [ ] Implement - [ ] Add to example - [ ] Add...

**Requirement Text** Better handling of dtype in data input. Sometimes its object- try converting to float **DoD** What need to be completed for this feature to be complete. E.g., -...

Add a link to the file in GitHub (master branch) in the doc comments. This way it will link in the generated documents

**Requirement Text** add penalty in loss function if outside range **Background Information** Outputs frequently have an expected range, and we want to encourage the model to learn something that will...

Example generating future loading schemes from planned activities. This is a more realistic scheme. Could be added to end of future loading example