DoE2Vec model that can transform any design of experiments (function landscape) to a feature vector.
For different input dimensions or sample size you require a different model.
Each model name is build up like doe2vec-d{dimension}-m{sample size}-ls{latent size}-{AE or VAE}-kl{Kl loss weight}
Example code of loading this huggingface model using the doe2vec package.
Then import and load the model.
1from doe2vec import doe_model
2
3obj = doe_model(
4 2,
5 8,
6 latent_dim=32,
7 kl_weight=0.001,
8 model_type="VAE"
9)
10obj.load_from_huggingface()
11#test the model
12obj.plot_label_clusters_bbob()
The model is intended to be used to generate feature representations for optimization function landscapes.
The representations can then be used for downstream tasks such as automatic optimization pipelines and meta-learning.
The model is trained using a weighed KL loss and mean squared error reconstruction loss.
The model is trained using 250.000 randomly generated functions (see the dataset) over 100 epochs.
