Announcements
- May 7, 2020 (7:46 PM)
- I spotted a small mistake in my solutions for HW #6, but am unable to scan in a corrected version. The value given for f_X(3|4) should be 27/55 instead of 25/55.
- May 7, 2020 (6:42 AM)
- I posted the solutions for HW #6 in the Week 14 section of the Blackboard site.
I'll try to e-mail you with info about HW #6 paper in the next day or two.
- May 6, 2020 (10:25 PM)
- Please enjoy this final exam ... it's due at 10:15 PM on Wed, May 13.
Link to the final exam.
- May 5, 2020 (4:26 PM)
- I created a
final exam review web page which includes a list of homework
parts to review, and some example problems based on the Ch. 10 material that will be covered during the last lecture.
- Apr. 23, 2020 (2:47 AM)
- I put link to the solutions for HW #5 in the Week 13 section of the class Blackboard site. There were 2 perfect scores of 36,
4 scores in the interval [30.6, 34.5] and 2 scores in the
interval [26.9, 27.5]. The median score is 33.25.
*** I think the recording I made of the lecture was successful. (I didn't listen to all of it, but the first part was good, and there is an indication that
it is about 2.5 hours long.) I haven't figured out how to put a link in Blackboard so that you can get to the recording easily, but I can explain how to listen to it.
First, click on the Week 12 tab, and then within Week 12 click on the link to Blackboard Collaborative Ultra. Near the left end of the black band going across (near
the top) with Sessions on it, click on the three horizonatal white line segments (the menu tab). Then click Recordings, and then click
Class 12, 652 / recording 1, and from there it should be obvious what to do. *** Some of you are still having trouble with using X vs. using x. In a likelihood function, a posterior pdf, or an estimate formula, you should use x_i (or just x if there is only one random variable). For an estimator you should use X_i.
Also, some of you do not seem to write X differently from x. When grading the final exam I'm going to be strict about the proper use of X vs. x, and if I cannot tell
whether you mean X or x you will not get the benefit of the doubt.
- Apr. 10, 2020 (6:17 PM)
- I put link to the solutions for HW #4 in the Week 11 section of the class Blackboard site. There were 4 scores in the interval [23.2, 28.8] and 4 scores in the
interval [34.6, 38.65]. The median score is 31.7 (out of 42 points possible, so about 75.5%).
*** I think the recording I made of the lecture was successful. (I didn't listen to all of it, but the first part was good, and there is an indication that
it is about 2.5 hours long.) I haven't figured out how to put a link in Blackboard so that you can get to the recording easily, but I can explain how to listen to it.
First, click on the Week 10 tab, and then within Week 10 click on the link to Blackboard Collaborative Ultra. Near the left end of the black band going across (near
the top) with Sessions on it, click on the three horizonatal white line segments (the menu tab). Then click Recordings, and then click
Class 10, 652 / recording 1, and from there it should be obvious what to do.
- Apr. 8, 2020 (12:22 PM)
- I put a link to HW #6 on the
homework web page.
- Mar. 26, 2020 (9:09 PM)
- I have scanned in your graded HW #3 papers and e-mailed them to you. (Unfortunately, I don't think I'll be able to do this for future assignments, but I
do plan to post detailed solutions each time. So keep a copy of your solutions for future assignments, and then you can compare your solutions to mine.)
I've put the solutions for HW #3 on the Blackboard site (look in the Week 9 area). Also I put a link to HW #5 on the
homework web page (of the course web site I used prior to doing some stuff on Blackboard). Even though I didn't cover as many
pages I intended to in our online class session last night, I did finish covering the pertinent material for HW #4. Next time I ought to cover all of the material
pertinent to Problem 2 of HW #5 and some of the material pertinent to the first problem of HW #5. *** I hope to be able to teach from my GMU office next
Wednesday, but then a week from today I plan to drive to Charleston, SC, and set up an office there (where I will be able to hook my GMU desk top computer up to the
internet via a cable, instead of having to rely on using Wi-Fi and a laptop from home .... I hope to have a printer with me, but I doubt that I'll have a scanner).
If I fail to show up on Blackboard next week or the one after that, assume that something about my plans got messed up. Instead of sending me a bunch of e-mail,
please just keep checking my old course web site and Blackboard for news from me ... I'll try to update you on the situation as soon as I can.
- Mar. 18, 2020 (9:24 PM)
- I updated the
syllabus
web page to reflect the revised schedule for the semester. (All dates for the 2nd portion of the semester have been shifted one week due
to the extension of Spring Break.) Also, it shows that the final exam will now be a take-home exam (that will be due on May 13).
- Mar. 17, 2020 (5:37 PM)
- I'll continue to use this web site to post announcements, homework assignments, and eventually a study guide for the final exam. So continue to check this web
site regularly. (I won't send e-mail every time I add something new here unless it's rather urgent.)
I'll use Blackboard for lectures, virtual office hours (perhaps), and the posting of lecture notes.
- Mar. 17, 2020 (4:25 PM)
- Our STAT 652 class resumes a week from tomorrow in online mode. Specifically, I'll be using Blackboard Collaborative Ultra to communicate with you.
Until this past Saturday, I had never used Blackboard before (since I prefer to make my own course web sites), and I've never taught online before. That being
the case, I'm going to follow the recommendations that faculty have been sent about the conversion to online teaching, and teach in a synchronous manner;
which means that instead of using prerecorded lectures, you will have to listen to me during the official 7:20 to 10:00 PM time slot on Wednesdays, starting next
week on the 25th. However, I'd like to do a quick test tomorrow at 7:20 PM, and I hope that some of you can join me in order to help me determine if things
are working as I expect them to. For this test, go to the STAT 652 Blackboard site and in the green panel on the left you should
see Week 8. Click there, and then click Blackboard Collaborative Ultra, and then click Test Session. (It's dated 3/15, because I first
created it an opened it on Sunday.) I'll be there by 7:10 in case you arrive early, and I'll stay there until 7:40. Please be patient since this will be my second
time connecting with anyone else using Collaborate Ultra. (Participation in the test session is optional, but I hope at least a couple of you can make it
in order to help me test my understanding of using Collaborative Ultra.) *** Note: On the 25th, you'll meet me the same way, except instead of clicking
on Test Session inside of Collaborative Ultra, you'll click on Class 8, STAT 652. On the 25th, HW #3 will be due (no lenghty grace period ... get
it to me by 11 PM on the 25th).
- Mar. 12, 2020 (1:36 AM)
- Late on Wednesday, the faculty was informed that Spring Break would be extended one week, and then all classes would be taught online (at least for a few weeks,
and perhaps longer). At this point I don't have enough information to allow me to make good decisions about how to handle everything, but I will state that
I've changed the due date of the next assignment to be submitted to now be Wednesday, the 25th.
I guess everyone will need to either fax or e-mail their solutions since the administration does not seem to want students coming to campus. If you can create a .pdf
file to attach to e-mail, that may be the best thing to do. If you scan handwritten solutions, be sure to write darkly so that the writing on the scanned version is
sufficiently dark. Do not e-mail photos of handwritten pages taken with a phone.
While we're in online mode, I've eliminated the grace period for homework (and so assignments will be due by 10 PM on the Wed. due date). *** In case some of you
have finished HW #3 and want to start on HW #4, I went ahead and posted it. Ordinarily, I would advise that you wait until I cover the UMVUE examples during the
8th lecture before attempting HW #4, but since having an online version of he 8th lecture isn't going to happen soon, some of you may want to carefully study all of
the UMVUE examples in the class notes and then start on the homework problems.
- Feb. 28, 2020 (3:16 AM)
- I finished grading HW #2. Most did pretty good on it, but lost some points due to not adequately justufying some answers. E.g., Problems 3 and 4 were fairly simple,
and one could perhaps make a correct guess about the limiting distribution. But still you were suppose to firmly establish your answer using methods covered in class,
and not simply indicate what seems to be suggested. *** At the end of the semester, the homework component of your final grade will be based on how many points
you earn compared to 90% of the total points that were available to be earned (not counting extra credit possibilities). There were 70 points available
on the first two assignments (plus 2 extra credit points), and so to judge how you're doing so far you can compare the sum of your first two HW scores to
(0.9)70 = 63. Written in red ink near your HW #2 score, I've given the percentage of the 63 points that you've earned. Four of you have a percentage of 92 or
higher, three of you have a percentage between 75 and 83, and one student has a percentage below 60. (Usually students getting an A+, A, or A- in STAT 652 have
homework percentages exceeding 90.)
- Feb. 26, 2020 (7:11 PM)
- I put a link to HW #3 on the
homework web page.
- Feb. 13, 2020 (6:00 AM)
- I finished grading HW #1, for which 32 was the highest possible score. 3 students got scores in the interval [31.5, 32], and
5 students got scores in the interval [22, 27]. The median of the scores is 26.6 (83.125%). (Based on just these scores, I'd say 3 grades of A, 4 Bs, and 1 B-.)
Some had trouble identifying the correct pdf to match an mgf, and some had trouble with properties of the natural log, and differentiation. Some perhaps should
have used a natural parameterization more.
- Jan. 30, 2020 (4:38 AM)
- I went ahead and posted a link to HW #2 on the
homework web page
even though it's based on material I'll cover during the next two lectures, and I think it'll be best to wait until after the next
lecture to start on it. (I'm worried that if I have an operation, it'll really hinder my ability to work.) *** I feel bad that I didn't get through p. 5.5.3 of the course
notes as I had planned to. Next time I will cover the bottom of p. 5.3.2, and then cover p. 5.3.3. I'll skip the 2 Sec. 5.4 pages, and then start on the lengthy
Sec. 5.5.
- Jan. 3, 2020 (6:28 PM)
- I just finished modifying this web site to make it ready for the start of the
Spring 2020 semester.
(If you visited the web site during the past 3 hours, you caught me
in the middle of modifying it.)
Please take time to read the
syllabus rather
carefully, and also the policies and comments about homework on
this web page.
I'll expect students to comply with all of the stated rules and
procedures.
*** I'll warn you up front that this is known to be one of the more difficult M.S.
level courses offered by my department, although my experience has been that students who do poorly are those who start the
semester without having a good enough background in calculus and probability. If you haven't sufficiently mastered the
prerequisite material, I advise you to wait until later, at some time when you have become better prepared, to take this course.
But to provide some evidence that the final exam will not be unreasonable, I'm leaving my final post to my students from the Spring 2015 semester below.
*** I went ahead and posted a link to the first homework assignment on the
homework web page.
The first three problems can be done using the material from Sec. 3.4 of the text that I'll cover during our first class meeting. The fourth problem is based on some
Ch. 5 material that I'll cover during our second class meeting.
(The first assignment isn't due until Sep. 17, but I strongly recommend that you try to solve the first three problems on it shortly after the first lecture.
I think that the first assignment is much simpler than the other 5 assignments will be, and if you're struggling to solve the first three problems on it, you should
perhaps interpret that as a sign that your calculus and probability skills are not strong enough for you to be successful in this class (and drop the class while you
still can without academic penalty). I'll also point out that a good portion of the first lecture will focus on methods that can be used to solve Problems 1-3. While it's possible
to also obtain the answers using standard probability methods covered in STAT 544, I recommend waiting until after the first lecture to start the homework so that you can perhaps
more comfortably use the special results for exponential families that I'll cover in class.)
- May 10, 2015 (4:24 AM)
- I have finished grading the exams, and now have the final course grades submitted to Patriot Web. On the whole, this semester's final exams
are the best I've ever gotten from STAT 652 students, with 6 scores in the 90s (high score: 97.45), and a median score of 81.375. When I computed the final
course averages for everyone, I noticed that they separated into distinct clusters, and so that's how I assigned the grades. There are large gaps between
the A, A-, B+, B, B-, C, and F clusters, and so no one just missed getting the next higher grade.