smpadhy/smart-ml-hpc-hug0

Model description

This is a Histogram-based Gradient Boosting Classification Tree model trained on HPC history jobs between 1Feb-1Aug 2022, window number 0.

Window Start: 2022-02-01 00:06:58; Window End: 2022-03-03 04:05:20; Total Jobs in Window 0: 35812.

Best parameters: {'hgbc__learning_rate': 0.1, 'hgbc__max_depth': 9, 'hgbc__max_iter': 600}

Performance on TEST

Accuracy on entire set: 0.946168166304685

Accuracy for last bin scheduling assuming bins <= 0 are incorrect: 0.9454; (936/990)

Accuracy for last bin scheduling assuming bins <= 1 are incorrect: 0.9242; (915/990)

Accuracy for last bin scheduling assuming bins <= 2 are incorrect: 0.9121; (903/990)

Accuracy for last bin scheduling assuming bins <= 3 are incorrect: 0.8878; (879/990)

Intended uses & limitations

[More Information Needed]

Training Procedure

[More Information Needed]

Hyperparameters

Click to expand

Hyperparameter	Value
memory
steps	[('scale', StandardScaler()), ('hgbc', HistGradientBoostingClassifier(max_depth=9, max_iter=600))]
verbose	False
scale	StandardScaler()
hgbc	HistGradientBoostingClassifier(max_depth=9, max_iter=600)
scale__copy	True
scale__with_mean	True
scale__with_std	True
hgbc__categorical_features
hgbc__class_weight
hgbc__early_stopping	auto
hgbc__interaction_cst
hgbc__l2_regularization	0.0
hgbc__learning_rate	0.1
hgbc__loss	log_loss
hgbc__max_bins	255
hgbc__max_depth	9
hgbc__max_iter	600
hgbc__max_leaf_nodes	31
hgbc__min_samples_leaf	20
hgbc__monotonic_cst
hgbc__n_iter_no_change	10
hgbc__random_state
hgbc__scoring	loss
hgbc__tol	1e-07
hgbc__validation_fraction	0.1
hgbc__verbose	0
hgbc__warm_start	False

Model Plot

Pipeline(steps=[('scale', StandardScaler()),('hgbc',HistGradientBoostingClassifier(max_depth=9, max_iter=600))])

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Evaluation Results

Metric	Value
accuracy	0.946168166304685
classification report	precision recall f1-score support 0 0.97 0.98 0.98 5075 1 0.74 0.57 0.64 218 2 0.70 0.59 0.64 108 3 0.67 0.55 0.60 86 4 0.89 0.92 0.90 959 accuracy 0.95 6446 macro avg 0.79 0.72 0.75 6446 weighted avg 0.94 0.95 0.94 6446

How to Get Started with the Model

[More Information Needed]

Model Card Authors

This model card is written by following authors:

[More Information Needed]

Model Card Contact

You can contact the model card authors through following channels: [More Information Needed]

Citation

Below you can find information related to citation.

BibTeX:

[More Information Needed]

citation_bibtex

bibtex @inproceedings{...,year={2024}}

get_started_code

import pickle with open(dtc_pkl_filename, 'rb') as file: clf = pickle.load(file)

model_card_authors

Smruti Padhy Joe Stubbs

limitations

This model is ready to be used in production.

model_description

This is a Histogram-based Gradient Boosting Classification Tree model trained on HPC history jobs between 1Feb-1Aug 2022, window number0

eval_method

The model is evaluated using test split, on accuracy and F1 score with macro average.

confusion_matrix