DeepTopPush: Simple and Scalable Method for Accuracy at the Top.
Year: 2020
Type: preprint
Abstract: Accuracy at the top is a special class of binary classification problems where the performance is evaluated only on a small number of relevant (top) samples. Applications include information retrieval systems or processes with manual (expensive) postprocessing. This leads to the minimization of irrelevant samples above a threshold. We consider classifiers in the form of an arbitrary (deep) network and propose a new method DeepTopPush for minimizing the top loss function. Since the threshold depends on all samples, the problem is non-decomposable. We modify the stochastic gradient descent to handle the non-decomposability in an end-to-end training manner and propose a way to estimate the threshold only from values on the current minibatch. We demonstrate the good performance of DeepTopPush on visual recognition datasets and on a real-world application of selecting a small number of molecules for further drug testing.
Source: arXiv (Cornell University)
Authors: Lukáš Adam, Václav Mácha, Václav Šmídl
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3036277013