Search

Summary

Previous chapters exclusively considered attacks against classifiers. In this chapter, we devise a backdoor attack and defense for deep regression or prediction models. Such models may be used to, for example, predict housing prices in an area given measured features, to estimate a city’s power consumption on a given day, or to price financial derivatives (where they replace complex equation solvers and vastly improve the speed of inference). The developed attack is made most effective by surrounding poisoned samples (with their mis-supervised target values) by clean samples, in order to localize the attack and thus make it evasive to detection. The developed defense involves the use of a kind of query-by-synthesis active learning which trades off depth (local error maximizers) and breadth of search. Both the developed attack and defense are evaluated for an application domain that involves the pricing of a simple (single barrier) financial option.

In this paper, the second order convergence of the interpolation based on -element is derived in the case of d=1, 2 and 3. Using the integral average on each element, the new basis functions of tensor product type is builded up and we can easily extend it to the higher dimensional case. Finally, some numerical tests are made to show the analytical results of the interpolation errors.

Search Results

Refine search

Refine search

Actions for selected content:

2 results

12 - Robust Deep Regression and Active Learning

Summary

Second Order Convergence of the Interpolation based on -Element

Search Results

Refine search

Refine search

Actions for selected content:

Save Search

2 results

12 - Robust Deep Regression and Active Learning

Summary

Second Order Convergence of the Interpolation based on -Element