Modeling Temporal Data as Continuous Functions with Stochastic Process Diffusion

Authors: Marin Biloš, Kashif Rasul, Anderson Schneider, Yuriy Nevmyvaka, Stephan Günnemann

ICML 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In Section 5 we empirically show that our model outperforms the baselines on all tasks. 5. Experiments
Researcher Affiliation Collaboration 1Technical University of Munich, Germany 2Machine Learning Research, Morgan Stanley, United States.
Pseudocode Yes In Algorithms 1 and 2 we provide the pseudocode for training the model and sampling new data, for DSPD-GP model. Algorithm 1 Training (DSPD-GP diffusion) Algorithm 2 Sampling (DSPD-GP diffusion)
Open Source Code Yes https://github.com/morganstanley/MSML/tree/main/ papers/Stochastic_Process_Diffusion
Open Datasets Yes We test our model as defined in Section 4.1 and Figure 2 against Time Grad (Rasul et al., 2021b) on three established real-world datasets: Electricity, Exchange and Solar (Lai et al., 2018). Table 4. Multivariate dimension, domain, frequency, total training time steps, and prediction length properties of the training datasets used in the forecasting experiments.
Dataset Splits No The paper uses "training data" and "test set" but does not provide specific split percentages, sample counts, or explicit cross-validation methodology for their experiments.
Hardware Specification No The paper does not provide specific hardware details such as GPU/CPU models, memory specifications, or cloud computing instance types used for the experiments.
Software Dependencies No The paper does not specify version numbers for any software dependencies, such as Python, PyTorch, or other libraries, that would be needed for replication.
Experiment Setup No The paper discusses aspects of model architecture and training process (e.g., Algorithms 1 and 2) but does not provide specific hyperparameter values (e.g., learning rate, batch size, number of epochs) or a dedicated section detailing the experimental setup with concrete training configurations in the main text.