The impact of validation set selected from test set
GaoYQ3991 opened this issue · 2 comments
GaoYQ3991 commented
Dear Ashley,
Thanks for your work fixing some common mistakes in other open source implementations. I found that the validation set was selected from the test set according to the proportion of 10%. Does this lead to an inaccurate test result?
ashleygritzman commented
In our final results we used the entire test set. Please go ahead and do some experiments if you think there may be any discrepancies.
Ashley Gritzman, PhD
Research Scientist
Phone: +27 (0)11 276 9236
Mobile: +27 (0)82 222 2852
E-mail: ashley.gritzman@za.ibm.com
Profile: http://ibmurl.hursley.ibm.com/NU58
IBM Research | South Africa Lab 47 Juta Street, Braamfontein, 2000 Johannesburg, South Africa
----- Original message -----From: GaoYQ3991 <notifications@github.com>To: IBM/matrix-capsules-with-em-routing <matrix-capsules-with-em-routing@noreply.github.com>Cc: Ashley Gritzman <ashley.gritzman@za.ibm.com>, Mention <mention@noreply.github.com>Subject: [EXTERNAL] [IBM/matrix-capsules-with-em-routing] The impact of validation set selected from test set (#5)Date: Thu, Nov 14, 2019 11:03 AM
Dear Ashley,Thanks for your work fixing some common mistakes in other open source implementations. I found that the validation set was selected from the test set according to the proportion of 10%. Does this lead to an inaccurate test result?
@ashleygritzman
—You are receiving this because you were mentioned.Reply to this email directly, view it on GitHub, or unsubscribe.
GaoYQ3991 commented
Thanks for your quick response. I will continue to study and experiment.