Deep imitation learning for molecular inverse problems
Authors: Eric Jonas
NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate on both synthetic and real data, showing that the resulting problem is invertible, is reasonably well-posed in the face of input data perturbations, and crucially performs well on experimentally-observed data for many molecules. For each spectrum in the test set our model generates 128 candidate structures and measure the fraction which are SMILES-correct. |
| Researcher Affiliation | Academia | Eric Jonas Department of Computer Science University of Chicago ericj@uchicago.edu |
| Pseudocode | No | The paper does not contain a dedicated 'Pseudocode' or 'Algorithm' section or figure. |
| Open Source Code | No | The paper does not provide an explicit statement or a link to open-source code for the methodology described. |
| Open Datasets | Yes | We obtain molecular structures consisting of H, C, O, and N atoms by downloading a large fraction of the public database Pub Chem []. ... We use NMRShift DB [18] for all data. |
| Dataset Splits | No | The paper states, 'This leaves us with 1,268,461 molecules, of which we train on 819,200 and draw the test set from the remainder.' It does not mention a separate validation set or specific validation split percentages/counts. |
| Hardware Specification | Yes | We train the network for 100 hours on an n Vidia v100 GPU. |
| Software Dependencies | No | The paper mentions software like 'Adam' (optimizer) and 'RDKit', but does not provide specific version numbers for these or other software dependencies. |
| Experiment Setup | Yes | We train using binary cross-entropy loss between the allowed possible next edges and the predicted edges of the network. We use Adam with a learning rate of 10 4 and train the network for 100 hours on an n Vidia v100 GPU. |