Answers to Assignments

Hi

Does anyone have the answers to all the tutorials?

I dont think anybody has the answers...
i havent received anything for assignments 2 and 3 yet

I have only received Assignment 1 back & marked... no answers to assignments, and no mark for assignment 2...

I got my marked Assignment 3 back yesterday... and i see the marks are up on MyUnisa... But the marked assignment didnt come with model answers... still waiting on that

Still haven't received Assignment 2 back, or a mark for it on MyUnisa.

check with the lecturer
i got my Assign 2 mark on MyUnisa, but the marked assignment wasnt posted to me.

Is there anyone else who didn't receive marks for an assignment(s) and didn't receive the marked assignment?

The lecturer remembers marking my assignment 2, but somehow the mark wasn't uploaded on myUnisa, and the marked assignment wasn't posted/got lost... there's no way I'm going to pass this course without that mark!

Any news on model answers to the assignments?

I still havent received model solutions... This is UNBELIEVABLE! Did anybody happen to receive solutions? Or at least heard from the lecturer?

i've received all my assignments back (and have points for all of them on myunisa), but i don't have any model answers unfortunately, don't think anyone has

i'm still struggling to get hold of the "Evaluation of outcome prediction for a clinical diabetes database" article. I sent the lecturer an email months ago asking if there was anywhere we could download it (without having to pay $25!), but received no answer.

seeing as how the exam is tomorrow, i'm probably going to have to raid my piggybank and just fork out for the article. needless to say, i'm completely unimpressed by this, especially since this is the 9th subject i've done and the first one where i've had to pay for an article.

hi. i have the article as a .pdf... if you give me your email address i will get someone to email it to you... let me know ok.

Hi Lois,

Could you please send me a PDF copy of the "Evaluation of outcome prediction for a clinical diabetes database" article to kbrat001@gmail.com

Thanks

Hi Lois,

Will you please send me a copy as well at ckucera@gmail.com .

Thank you

hi there. i have sent the article to you guys. sorry about the delay - its hard accessing the internet via my home computer, and i generally use my phone to access the forums. anyway all the best for tomorrow.

Thank you very much and good luck to you too.

Hi

Thanks a mil for the offer, but I ended up just buying this article

I worked through Assignment 3, Question 6 - I think the answers should be something like the following (if anyone sees anthing drastically wrong here, feel free to shout!):

6.1
sensitivity = the fraction of true positive results that are correctly classified as positive
specificity = the fraction of true negative results that are correctly classified as negative

6.2.1
The three algorithms are:
- Naive Boyes
- IB1
- C 4.5 Decision Trees

Decision trees had the best overall accuracy.

6.2.2
The Naive Boyes algorithm had the best performance when attempting to identify good blood glucoce patients (highest sensitivity).

6.2.3
The Decision tree had the best performance when attempting to identify bad blood glucoce patients (highest specificity) and it also had the best overall Accuracy (a measurement calculated by combining specificity and sensitivity). Naive Bayes was less Accurate, but had the best capacity to keep the balance between sensitivity and specificity.

Best of luck to all of you for tomorrow!

Hi there. With regards to the answers. Mine was - 6.2.2. Discretized C4.5 - 6.2.3. Naive Bayes. And these were marked right. This is how i understand it: high sensitivity = good performance for pos control, and similarly for specificity.

And small difference between sensitivity and specifity = performs well when both pos and neg control is considered. the values of the last row (average) in Table 2 show this. hope it helps. thanks.

oh yes i think confusion arises because in this article, positive control is the bad blood glucose cases and neg control is the good blood glucose cases - bottom of pg 185

Guys,thanks for sharing the article and your views on the solutions .It helps to have something to compare against.Anyone care to share some guidance on Assignment 2 question 3.2?I've really battled with it

i havent received my marked assign2 but i think for the manual seq alignment there are many possible solutions. i jus introduced gaps in various places and tried to get a better hit count

Good luck everybody! its a two-hour paper so lets hope its short and easy.