US 11,995,547 B2
Predicting and visualizing outcomes using a time-aware recurrent neural network
Fan Du, Milpitas, CA (US); Sungchul Kim, San Jose, CA (US); Shunan Guo, San Jose, CA (US); Sana Lee, ShangHai (CN); and Eunyee Koh, Cupertino, CA (US)
Assigned to ADOBE INC., San Jose, CA (US)
Filed by Adobe Inc., San Jose, CA (US)
Filed on Aug. 30, 2022, as Appl. No. 17/823,390.
Application 17/823,390 is a continuation of application No. 16/394,227, filed on Apr. 25, 2019, granted, now 11,475,295.
Prior Publication US 2022/0414468 A1, Dec. 29, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 3/08 (2023.01); G06N 3/04 (2023.01); G06N 3/045 (2023.01); G06N 5/02 (2023.01); G06N 7/01 (2023.01); G06N 20/10 (2019.01)
CPC G06N 3/08 (2013.01) [G06N 3/045 (2023.01); G06N 5/02 (2013.01); G06N 7/01 (2023.01); G06N 20/10 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
encoding a sequence of events into a feature vector comprising, for each event in the sequence of events, a numerical representation of (i) a respective category of the event and (ii) a respective timestamp of the event;
applying, to the feature vector, a network that outputs a sequence embedding comprising a probability distribution of a plurality of future events and an associated duration for each future event, wherein the network accommodates time irregularities in the sequence of events;
applying, to the sequence embedding, a classifier that computes a likelihood of a categorical outcome for each of the events in the probability distribution;
determining, via the classifier, one or more additional events that, if added to the sequence of events, would result in a categorical outcome matching a user-specified category; and
providing one or more of (i) the probability distribution, (ii) the categorical outcome or (iii) the additional events to a user interface.