Timezone: »

 
Spotlight
Deep Probability Estimation
Sheng Liu · Aakash Kaku · Weicheng Zhu · Matan Leibovich · Sreyas Mohan · Boyang Yu · Haoxiang Huang · Laure Zanna · Narges Razavian · Jonathan Niles-Weed · Carlos Fernandez-Granda

Tue Jul 19 01:25 PM -- 01:30 PM (PDT) @ None

Reliable probability estimation is of crucial importance in many real-world applications where there is inherent uncertainty, such as weather forecasting, medical prognosis, or collision avoidance in autonomous vehicles. Probability-estimation models are trained on observed outcomes (e.g. whether it has rained or not, or whether a patient has died or not), because the ground-truth probabilities of the events of interest are typically unknown. The problem is therefore analogous to binary classification, with the important difference that the objective is to estimate probabilities rather than predicting the specific outcome. The goal of this work is to investigate probability estimation from high-dimensional data using deep neural networks. There exist several methods to improve the probabilities generated by these models but they mostly focus on classification problems where the probabilities are related to model uncertainty. In the case of problems with inherent uncertainty, it is challenging to evaluate performance without access to ground-truth probabilities. To address this, we build a synthetic dataset to study and compare different computable metrics. We evaluate existing methods on the synthetic data as well as on three real-world probability estimation tasks, all of which involve inherent uncertainty: precipitation forecasting from radar images, predicting cancer patient survival from histopathology images, and predicting car crashes from dashcam videos. Finally, we also propose a new method for probability estimation using neural networks, which modifies the training process to promote output probabilities that are consistent with empirical probabilities computed from the data. The method outperforms existing approaches on most metrics on the simulated as well as real-world data.

Author Information

Sheng Liu (NYU)
Aakash Kaku (New York University)
Weicheng Zhu (New York University)
Matan Leibovich (New York University)
Sreyas Mohan (NYU)
Boyang Yu (NYU Center for Data Science)
Haoxiang Huang (New York University)
Laure Zanna (NYU)
Narges Razavian (New York University)
Jonathan Niles-Weed (NYU)
Carlos Fernandez-Granda

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors