Human-Centric Justification of Machine Learning Predictions
Authors: Or Biran, Kathleen McKeown
IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Through a taskbased experiment, we show that our approach significantly helps humans to correctly decide whether or not predictions are accurate, and significantly increases their satisfaction with the justification. |
| Researcher Affiliation | Collaboration | n-Join or@n-join.com Kathleen Mc Keown Columbia University kathy@cs.columbia.edu |
| Pseudocode | No | The paper describes methods and processes in text, including mathematical formulations for linear classifiers, but it does not present any structured pseudocode blocks or algorithms. |
| Open Source Code | Yes | We made our justification method publicly available as a Java package.2 http://www.cs.columbia.edu/ orb/preju/ |
| Open Datasets | Yes | The classifier uses the 23 features shown in Table 2 and is trained on two years of daily pricing data for the S&P500 companies, available on Yahoo! Finance. |
| Dataset Splits | No | The paper mentions training and testing data but does not specify a separate validation split, its size, or how it was used in the experimental setup for reproducibility. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used for running the experiments (e.g., CPU, GPU models, memory, or cloud instance types). |
| Software Dependencies | No | The paper mentions 'Java package' and 'Weka’s Logistic Regression and SMO classifiers' but does not specify version numbers for Java, Weka, or any other software libraries, which is necessary for reproducibility. |
| Experiment Setup | No | The paper states that a Logistic Regression classifier was used with 23 features and trained on two years of data, but it does not provide specific experimental setup details such as hyperparameters (e.g., learning rate, batch size, number of epochs, optimizer settings) needed to reproduce the training of the model. |