Computer-implemented methods and nodes implementing performance estimation of algorithms during evaluation of data sets using multiparty computation based random forest
AssigneeKoninklijke Philips N.V.
InventorsLu Zhang, Meilof Geert Veeningen
Life of this patent
Estimating the term…
Abstract
According to an aspect, there is provided a computer-implemented method of operating a first node. The first node has an algorithm for evaluating input data from another node, with the input data having a plurality of different attributes. The method comprises receiving, from a second node, a proposal for the evaluation of a first set of input data by the algorithm; estimating the performance of the algorithm in evaluating the first set of input data based on the proposal; and outputting, to the second node, an indication of the estimated performance of the algorithm. A corresponding first node is also provided.
Claims24
A computer-implemented method of operating a first node, the first node having an algorithm for evaluating input data from another node, the input data having a plurality of different attributes, the method comprising:
- receiving, from a second node having access to one or more data sets that are private to the second node, a proposal for the evaluation of a first set of input data by the algorithm, wherein the first set of input data comprises attributes that are private to the second node and wherein the proposal does not disclose the first set of input data to first node;
- estimating the performance of the algorithm in evaluating the first set of input data based on the proposal, wherein any part of the first set of input data for which attributes are private to the second node is to be evaluated using a multiparty computation, MPC-based random forest, and wherein the step of estimating the performance of the algorithm comprises determining an indication of a time required to complete the evaluation of the first set of input data by:
- determining an evaluation time cost for evaluating all encrypted trees in the random forest to be used for the evaluation of the first set of input data, wherein the encrypted trees are used to evaluate attributes of the first set of input data that are private to the second node;
- determining an input time cost for evaluating all inputs to the algorithm, wherein the input time cost is the time cost of inputting private attributes to the algorithm;
- and
- determining the indication of the time required as the sum of the determined evaluation time cost and the determined input time cost;
- and
- outputting, to the second node, an indication of the estimated performance of the algorithm;
- receiving, from the second node, a second indication that the estimated performance is acceptable;
- and
- evaluating, by the first node, at least a portion of the first set of input data using the algorithm in response to receiving the second indication.
A first node, the first node having an algorithm for evaluating input data from another node, the input data having a plurality of different attributes, wherein the first node is configured to:
- receive, from a second node having access to one or more data sets that are private to the second node, a proposal for the evaluation of a first set of input data by the algorithm, wherein the first set of input data comprises attributes that are private to the second node and wherein the proposal does not disclose the first set of input data to the first node;
- estimate the performance of the algorithm in evaluating the first set of input data based on the proposal, wherein any part of the first set of input data for which attributes are private to the second node is to be evaluated using a multiparty computation, MPC-based random forest, and wherein the step of estimating the performance of the algorithm comprises determining an indication of a time required to complete the evaluation of the first set of input data by:
- determining an evaluation time cost for evaluating all encrypted trees in the random forest to be used for the evaluation of the first set of input data, wherein the encrypted trees are used to evaluate attributes of the first set of input data that are private to the second node;
- determining an input time cost for evaluating all inputs to the algorithm, wherein the input time cost is the time cost of inputting private attributes to the algorithm;
- and
- determining the indication of the time required as the sum of the determined evaluation time cost and the determined input time cost;
- and
- output, to the second node, an indication of the estimated performance of the algorithm;
- receive, from the second node, a second indication that the estimated performance is acceptable;
- and
- evaluate at least a portion of the first set of input data using the algorithm in response to receiving the second indication.
Dependencies are inferred from the claim text.