Towards Better Modeling with Missing Data: A Contrastive Learning-based Visual Analytics Perspective

Laixin Xie, Yang Ouyang, Longfei Chen, Ziming Wu, Quan Li

Room: 109

2023-10-26T03:24:00ZGMT-0600Change your timezone on the schedule page
Exemplar figure, described by caption below
Interaction and observation in CIVis. The notation 1 - 3 demonstrate how to configure positive and negative sampling based on visual cues; notation 4 - 7 highlight and explain the benefits that CIVis brings. The loss curves in (4) look like an area because the loss sharply and frequently jumps up and down.
Fast forward
Full Video

Contrastive learning;data imputation;explainable AI;missing data


Missing data can pose a challenge for machine learning (ML) modeling. To address this, current approaches are categorized into feature imputation and label prediction and are primarily focused on handling missing data to enhance ML performance. These approaches rely on the observed data to estimate the missing values and therefore encounter three main shortcomings in imputation, including the need for different imputation methods for various missing data mechanisms, heavy dependence on the assumption of data distribution, and potential introduction of bias. This study proposes a Contrastive Learning (CL) framework to model observed data with missing values, where the ML model learns the similarity between an incomplete sample and its complete counterpart and the dissimilarity between other samples. Our proposed approach demonstrates the advantages of CL without requiring any imputation. To enhance interpretability, we introduce CIVis, a visual analytics system that incorporates interpretable techniques to visualize the learning process and diagnose the model status. Users can leverage their domain knowledge through interactive sampling to identify negative and positive pairs in CL. The output of CIVis is an optimized model that takes specified features and predicts downstream tasks. We provide two usage scenarios in regression and classification tasks and conduct quantitative experiments, expert interviews, and a qualitative user study to demonstrate the effectiveness of our approach. In short, this study offers a valuable contribution to addressing the challenges associated with ML modeling in the presence of missing data by providing a practical solution that achieves high predictive accuracy and model interpretability.