We display full results for all methods here. This means that the source feature quality doesn't matter nearly as much. Since source feature quality is the only metric these methods use to predict transfer performance, they do poorly here. In Tab. 1, we display the Pearson Correlation for each target dataset individually. We also include results for additional baselines and skews of existing methods.

Meta-learning is a promising strategy for learning to efficiently learn using data gathered from a distribution of tasks.

COOT leads to performance improvements over the state-of-the-art methods.

A.2 Proof of Theorem 1 To prove Theorem 1, we assume that G Proof of Lemma 1. Let's first rewrite Equation (4) as null null By Lemma 1, linearity of expectation and knowing that each RWT is independent from the other tours by the Strong Markov Property, Theorem 1 holds. MHM-GNN can recover edge-based models where representations don't use graph-wide However, on Rent the Runway we see the raw features achieving the highest performance. That is, structural information does not seem to be relevant to this specific task. All hyperparameters were chosen to minimize training loss. For k = 5, we used a minibatch of size 5 in all datasets.

Figure 1: The proposed unsupervised graph representation using motif compositions.

Cross-Domain Recommendation (CDR) has been popularly studied to utilize different domain knowledge to solve the cold-start problem in recommender systems.

In the main text, many algorithmic details were omitted and only discussed briefly. A.1 Dataset Details We expand upon the seven datasets used for our experiments in this section. The task is multi-class classification with a heavy class imbalance. It has 8 features including price, day of the week and units transferred. We discard instances with missing values.