Deep Recurrent Neural Networks (RNN) is increasingly used in decision-making with temporal sequences. However, understanding how RNN models produce final predictions remains a major challenge. Existing work on interpreting RNN models for sequence predictions often focuses on explaining predictions for individual data instances (e.g., patients or students). Because state-of-the-art predictive models are formed with millions of parameters optimized over millions of instances, explaining predictions for single data instances can easily miss a bigger picture. Besides, many outperforming RNN models use multi-hot encoding to represent the presence/absence of features, where the interpretability of feature value attribution is missing. We present ViSFA, an interactive system that visually summarizes feature attribution over time for different feature values. ViSFA scales to large data such as the MIMIC dataset containing the electronic health records of 1.2 million high-dimensional temporal events. We demonstrate that ViSFA can help us reason RNN prediction and uncover insights from data by distilling complex attribution into compact and easy-to-interpret visualizations.

Deep Recurrent Neural Network (RNN) has gained popularity in many sequence classification tasks. Beyond predicting a correct class for each data instance, data scientists also want to understand what differentiating factors in the data have contributed to the classification during the learning process. We present a visual analytics approach to facilitate this task by revealing the RNN attention for all data instances, their temporal positions in the sequences, and the attribution of variables at each value level. We demonstrate with real-world datasets that our approach can help data scientists to understand such dynamics in deep RNNs from the training results, hence guiding their modeling process.

Validation and exploration of the data generated by large-scale scientific simulations rely on sophisticated visualization and analysis tasks. With the advancement of supercomputing, the growing scale and complexity of the data make some of these tasks challenging, which demands new hardware and software solutions. We believe it is possible to address some of the challenges by utilizing the increasingly affordable see-through head mounted display (H-MD) devices together with a low-cost tiled HDTV display. With the tiled display to provide a high-resolution overview of the data, the user can freely choose a small area to explore and analyze using a see-through HMD in stereoscopic 3D with gesture input. During such local exploration and detail data analysis, the user can apply a newly derived visualization parameter setting to the large tiled display for a new overview. In this way, computational costs become more manageable because realtime rendering and response are only required to cover a small screen space and a subset of the data. In our current study, we focus on supporting immersive isosurface and streamline visualization and analysis of 3D flow field data. In this workshop paper, we present our preliminary design and results, and we also discuss our further development and evaluation plan.