plot accuracy & time vs. iteration #

[jovo]

for accuracy, use accuracy of both E step and M step
(don't count computing accuracy in the time calculation)

@yangyutu

[jovo]

how is this coming along?
@vincentliuk @yangyutu

once this is done, let's chat again?

[yangyutu]

Hello, Josh,

I am getting confused. Are we supposed to make a new figure or update some
previous figure?

Thanks.

On Tue, Aug 25, 2015 at 12:09 AM, joshua vogelstein <
notifications@github.com> wrote:

how is this coming along?
@vincentliuk https://github.com/vincentliuk @yangyutu
https://github.com/yangyutu

once this is done, let's chat again?

—
Reply to this email directly or view it on GitHub
#74 (comment).

[jovo]

new fig, 3 lines:
parameter accuracy vs. iteration
time vs. iteration
prediction accuracy vs. iteration.

now that we understand time within iterations,
we want to understand accuracy and time across iterations.

note: please include iteration 0, ie, the initialization phase.

is that clear?

On Tue, Aug 25, 2015 at 10:07 AM, Yuguang Yang notifications@github.com
wrote:

Hello, Josh,

I am getting confused. Are we supposed to make a new figure or update some
previous figure?

Thanks.

On Tue, Aug 25, 2015 at 12:09 AM, joshua vogelstein <
notifications@github.com> wrote:

how is this coming along?
@vincentliuk https://github.com/vincentliuk @yangyutu
https://github.com/yangyutu

once this is done, let's chat again?

—
Reply to this email directly or view it on GitHub
#74 (comment).

—
Reply to this email directly or view it on GitHub
#74 (comment).

the glass is all full: half water, half air.
openconnecto.me, jovo.me, office hours
https://www.google.com/calendar/embed?src=vk2j3ugqcitvq223g7ivvq4hqo%40group.calendar.google.com&ctz=America/New_York

[yangyutu]

Hello, Josh, I have three questions:
(1) time means computational time?
(2) you said "use accuracy of both E step and M step", then for each iteration, I need to plot the accuracy for both E step and M step? Then there will be 5 lines rather than 3.
(3) which data shall I use? I am planning to use the low dimensional setting as in Figure 1 for this plot.

[jovo]

yes.
ok
good start.
thanks.

On Tue, Aug 25, 2015 at 11:39 AM, Yuguang Yang notifications@github.com
wrote:

Hello, Josh, I have three questions:
(1) time means computational time?
(2) you said "use accuracy of both E step and M step", then for each
iteration, I need to plot the accuracy for both E step and M step? Then
there will be 5 lines rather than 3.
(3) which data shall I use? I am planning to use the low dimensional
setting as in Figure 1 for this plot.

—
Reply to this email directly or view it on GitHub
#74 (comment).

the glass is all full: half water, half air.
openconnecto.me, jovo.me, office hours
https://www.google.com/calendar/embed?src=vk2j3ugqcitvq223g7ivvq4hqo%40group.calendar.google.com&ctz=America/New_York