Skip to main content

Compares weaknesses in multiple models

Project description

AutoRA Model Disagreement Sampler

The model disagreement sampler identifies experimental conditions $\vec{x}' \in X'$ with respect to a pairwise distance metric between theorist models, $P_{M_{i}}(\hat{y}, \vec{x}')$:

$$ \underset{\vec{x}'}{\arg\max}~(P_{M_{1}}(\hat{y}, \vec{x}') - P_{M_{2}}(\hat{y}, \vec{x}'))^2 $$

Example Code

from autora.experimentalist.sampler.model_disagreement import model_disagreement_sample
from autora.theorist.bms import BMSRegressor; BMSRegressor()
from autora.theorist.darts import DARTSRegressor; DARTSRegressor()
import numpy as np

#Meta-Setup
X = np.linspace(start=-3, stop=6, num=10).reshape(-1, 1)
y = (X**2).reshape(-1, 1)
n = 5

#Theorists
bms_theorist = BMSRegressor()
darts_theorist = DARTSRegressor()
bms_theorist.fit(X,y)
darts_theorist.fit(X,y)

#Sampler
X_new = model_disagreement_sample(X, [bms_theorist, darts_theorist], n)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

Built Distribution

File details

Details for the file autora-experimentalist-sampler-model-disagreement-1.0.4.tar.gz.

File metadata

File hashes

Hashes for autora-experimentalist-sampler-model-disagreement-1.0.4.tar.gz
Algorithm Hash digest
SHA256 86493c70633aad10772be59cd4a299f12a510a26972fbf68449a5434436f1b5f
MD5 f6b0c274ba664336bf4c359cedb7fe61
BLAKE2b-256 1f50d2feedaa7e97a0c1b6ad33bcf7ae0c910d2d6f15ced841f6459004ef0d33

See more details on using hashes here.

File details

Details for the file autora_experimentalist_sampler_model_disagreement-1.0.4-py3-none-any.whl.

File metadata

File hashes

Hashes for autora_experimentalist_sampler_model_disagreement-1.0.4-py3-none-any.whl
Algorithm Hash digest
SHA256 976e9cdafb95d9b92ee2c19862f5fa730597efce115f3d5aa7da5dd696d1db1d
MD5 6016fa9badf533803faf2d337f2d8e56
BLAKE2b-256 afe8249143a7f987ab9cbcbc5b2ffcb14ea9290b1c6df6e84d0f986d899121da

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page