Timezone: »

How Can I Explain This to You? An Empirical Study of Deep Neural Network Explanation Methods
Jeya Vikranth Jeyakumar · Joseph Noor · Yu-Hsi Cheng · Luis Garcia · Mani Srivastava

Wed Dec 09 09:00 PM -- 11:00 PM (PST) @ Poster Session 4 #1172

Explaining the inner workings of deep neural network models have received considerable attention in recent years. Researchers have attempted to provide human parseable explanations justifying why a model performed a specific classification. Although many of these toolkits are available for use, it is unclear which style of explanation is preferred by end-users, thereby demanding investigation. We performed a cross-analysis Amazon Mechanical Turk study comparing the popular state-of-the-art explanation methods to empirically determine which are better in explaining model decisions. The participants were asked to compare explanation methods across applications spanning image, text, audio, and sensory domains. Among the surveyed methods, explanation-by-example was preferred in all domains except text sentiment classification, where LIME's method of annotating input text was preferred. We highlight qualitative aspects of employing the studied explainability methods and conclude with implications for researchers and engineers that seek to incorporate explanations into user-facing deployments.

Author Information

Jeya Vikranth Jeyakumar (University of California, Los Angeles)

I'm pursuing my Ph.D. in Electrical and Computer Engineering at UCLA, advised by professor Mani Srivastava. My current focus is on Interpretable deep learning for multi-modal sensor systems(cameras, sensors on smartphones and wearables, IoT devices, Bluetooth beacons, acoustic signals, etc) to infer human activities and biomarkers. My research interests include Machine Learning for IoT, Sensor Information Processing and Cyber-Physical systems. Besides research, I enjoy hiking, traveling (in general), science-fiction and get way too much happiness from good food!

Joseph Noor (University of California, Los Angeles)
Yu-Hsi Cheng (UCLA)
Luis Garcia (University of California, Los Angeles)
Mani Srivastava (UCLA)