AIConfPaper
← Search

Komminist Weldemariam

4 accepted papers

2020

Decision Platform for Pattern Discovery and Causal Effect Estimation in Contraceptive Discontinuation

Celia Cintas, Ramya Raghavendra, Victor Akinwande, Aisha Walcott-Bryant, Charity Wayua, Komminist Weldemariam
IJCAI 2020poster

Contraceptive use improves the health of women and children in several ways, yet data shows high rates of discontinuation which is not well understood. We introduce an AI-based decision platform capable of analyzing event data to identify patterns of contraceptive uptake that are unique to a subpopu…

Cited by 0Source ↗PDF ↗Scholar ↗
2020

Detecting Adversarial Attacks via Subset Scanning of Autoencoder Activations and Reconstruction Error

Celia Cintas, Skyler Speakman, Victor Akinwande, William Ogallo, Komminist Weldemariam, Srihari Sridharan +1
IJCAI 2020poster

Reliably detecting attacks in a given set of inputs is of high practical relevance because of the vulnerability of neural networks to adversarial examples. These altered inputs create a security risk in applications with real-world consequences, such as self-driving cars, robotics and financial ser…

Cited by 0Source ↗PDF ↗Scholar ↗
2020

Inspection of Blackbox Models for Evaluating Vulnerability in Maternal, Newborn, and Child Health

William Ogallo, Skyler Speakman, Victor Akinwande, Kush R Varshney, Aisha Walcott-Bryant, Charity Wayua +1
IJCAI 2020poster

Improving maternal, newborn, and child health (MNCH) outcomes is a critical target for global sustainable development. Our research is centered on building predictive models, evaluating their interpretability, and generating actionable insights about the markers (features) and triggers (events) asso…

Cited by 0Source ↗PDF ↗Scholar ↗
2020

Preservation of Anomalous Subgroups On Variational Autoencoder Transformed Data

Samuel C. Maina, Reginald E. Bryant, William O. Ogallo, Kush R. Varshney, Skyler Speakman, Celia Cintas +3
ICASSP 2020accepted

We investigate the effect of variational autoencoder (VAE) based data anonymization and its ability to preserve anomalous subgroup properties. We present a Utility Guaranteed Deep Privacy (UGDP) system which casts existing anomalous pattern detection methods as a new utility measure for data synthes…

Cited by 0Source ↗Scholar ↗