I need two versions one for me and one for my friends (each one 100 words) we are in the same class and same teacher. I attached some examples.I attached my classmate discussion and it has many reply sample Please let me know which two you are replying
week_10_discu.docx
Unformatted Attachment Preview
Week 10 – Discussion
Either describe one concept or skill from the course (especially week 9) that you think you
are well prepared to demonstrate on the final, including what helped you wrap your head
around it, OR describe one topic from the course (especially week 9) that you find
challenging, and ask for help.
(1)
One concept from chapter 9 that I feel I am well prepared to demonstrate on the final exam
is the logistic model. I hadn’t really looked at natural logs or ‘e’ (Euler’s number,
approximated at 2.718) since taking a calculus course about two years ago, but the ideas were
still somewhat familiar.
Similar to how linear regressions predict a dependent variable using the linear model (noted
in the visual above), logistic regressions are used in a similar capacity to predict binary
outcomes. The model produces an S-shaped curve that never goes below zero or above
one. This non-linear occurrence is a perfect application for 1) categorical data with a binary
result, and 2) probability predictions since values are measured between zero and one.
With respect to wrapping my heard around this concept, my first step was further building
my comfort and familiarity with the basic linear regression model as it shares some
characteristics with the logistic function. Next, I used the exercises in our book to recognize
the appropriate scenarios that warrant a logistic function rather than a linear one. Finally,
using software like Excel to find the estimated probability of an event with respect to certain
explanatory variables really helps; these calculations would be pretty tedious to do
manually. Page 452 of the textbook has an interesting problem related to logistic regression
for an online dating website.
Gerald,
What a great post. This was exactly what I needed to read, this has been one of the most
complicated topics for me so far I just did no really understand it. I get how important it can
be for probability predictions since all the numbers have to be between 0 and 1. I will try
myself the problem on pg 452, it might be very helpful for me.
Gerald,
Great post, when I first started working on this weeks project I didn’t quite understand what
the logistic model was. Your explanation with how the binary data numbers function within
the graph really help me wrap my head around this topic.
Thank you.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
(2)
I’m think I’ve got a pretty decent handle on simple linear regression and time series analysis,
though I’m easily confused when looking at the formulas in the text and I’m going to have to
take some time simplifying them into terms a non-mathematician such as myself can
comprehend prior to taking the final exam.
It’s comforting to know that the final exam is weighted so heavily that the B I’ve managed to
pull out of this class so far is completely irrelevant.
Tristan,
I have a B in this class too but I am really concerned with the fact that my exam grade can
bring it down a lot since it is so much percentage of the final exam. Do you have any tips on
studying or something special you might do in order to get yourself to a passing grade?
Hi Tristan,
Yes, I admit that some of the formulas can be a bit intimidating at first glance. I try to dissect
and deconstruct them in order to better understand the steps and procedures. For example,
when you look at the formula for MSE (mean squared error) on page 384, it looks rather
complicated. However, when you set up a simple spreadsheet in Excel to help with the
calculation, you realize that it’s only a few basic steps. Since we’re allowed to use a note page
for the final exam, I’m thinking it may be helpful to write up a basic spreadsheet with a few
entries so you can see how to compute some of the metrics (MAE, MSE, MFE, etc.)
It’s also worth noting that there’s a practice exam with an answer key in the Modules tab; I
think that’s a really helpful tool.
Tristan,
I feel pretty confident, but I struggle to keep track of formulas at times too. Frequently I’ve
found myself using the wrong formula, but as soon as I straighten that out I breeze through
the problem. I think the key to this exam is going to be really good notes, and closely reading
every problem.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
(3)
Over the entire course the skills I feel most confidant with is time series analysis and linear
regression. These make sense to me and have a limited number of variables to them. I don’t
think they are the easiest thing to do, but I think that looking at time series with exponential
smoothing is a very cool concept. The last project we did I felt very comfortable with and
created a really interesting graph.
If any one has questions about these concepts please post below to discuss.
Hi Patrick,
At first I found this challenging, but then I thought about in a logical sense and I got it. You
are right there are limited variables, so it’s easier to understand the concepts. Exponential
smoothing is also a cool concept, agreed! Graphically it can tell the story much easier.
Do you think a time series graph is best to have the exponential smoothing?
Patrick,
Do you have any tricks for remembering how to do linear regression? I was just going over
the homework and I think that I struggled with that the most overall.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
(4)
A concept that has been confusing for me of last week was the logistic model. I have had a
hard time trying to understand why the data in the graph moves similarly to when using the
linear model. The fact that it is an S shaped curve that can never go below 0 is hard for me to
wrap my head around. Is there a way to use excel in a simple way so that I can figure out
how to understand these?
I am actually really worried about the final exam because I have a pretty decent grade in the
class right now but I’m worried since I am not a good test taker. Does anyone have any tips
that could help me in order to at least get a passing grade in the final? Maybe someone that
has taken the class before and that knows that to expect?
Hey there, Daniela.
With respect to the logistic models, I personally prefer using Statgraphics rather than Excel
for that particular function. Excel is a very useful and valuable tool when it comes to many
applications, but I think Statgraphics is the superior choice for this form of analysis.
Once you upload your data into the program, go to Relate > Attribute Data > Logistic
Regression in the menu ribbon. From there, you can just use the default settings (preselected by the program) to display the output. With respect to your concern, the fact that a
logistic curve never dips below zero is simply a reference to probabilities (or a binary
outcome), which range from 0% – 100%. The curve doesn’t dip below zero because that is
the lowest possible probability — you can’t have a negative probability of an event
occurring. For binary outcomes, it’s either one or the other, A or B… the basketball team
either won or lost, you either own a dog or you don’t, you either have siblings or you don’t,
etc. Have a look at the Oscar nominations example on page 434 for additional color on this
topic.
With respect to the final exam, it’s my first time taking the class, so I’m a little anxious
too. However, the study guide and tools in Canvas are very thorough and comprehensive. I
feel confident that if you can master the practice exam in Canvas, you’ve probably got a
really good chance of doing well on the exam.
Hi Daniela,
I agree with you I am still struggling to completely understand the logistic model. I think
using excel to figure this out is a great idea! If anyone has further suggestions to better
understand this please share!
I am also not the best test taker, and seem to be in a similar position to you! For me it helps
when I study in small stints, rather than all at once the day before. I am going to look back
on all the projects as well as do some of the practice problems in cengage.
I would also love some advice!
Good luck, I’m sure you will do great.
Daniela,
I agree I am having trouble with the logistic model. I am having a hard time understanding
and visualizing it and getting it in my head to be able to perfectly use it on the test. I agree I
am very nervous about the test as well. I am a horrible test taker so I have been pretty
nervous. I have been studying in smaller amounts for a while now to hopefully have it
engrained in my head. I am nervous that I am going to stress and forget everything! I do
recommend going over the practice exams to see if you thoroughly understand the concepts
and also going to the stats tutors help. I do not have the best grade in the class but have been
trying my best to understand them to the best of my ability. Good luck!!
Daniela,
I’m in the same boat. The logistic model has not made the most sense to me, but some posts
here have helped a bit. As Gerald mentioned, it is easier to see in Statgraphics than Excel.
I am worried about the final as well. While I feel pretty confident on most of the topics, tests
are not my strong suite either and I tend to misread things and struggle to find the right
formula for the problem. I think my success on this test will largely depend on how good my
notes are and how much time I spend with the study materials.
…
Purchase answer to see full
attachment
