Q $1$: Homework Consolidation $200$ pts

Write a program to consolidate ALL your homeworks done to date $($leave out the LinkedIN one$).$

In essence, you should have $10$ homeworks. The user should be able to enter the homework number,

and the homework runs as it is expected to$.$ Once the homework run is complete, we should come back

to the original menu, where the user can select another homework $($or type $0$ to exit$).$

Note: Do Error Handling in menu inputs, e$.$g$.$ not allowing any input other than $0-10.$

Sample Menu

$1$: BMI

$2$: While Question String

$3$: FindRemainder Function

$4$: Lists with $0$

$5$: First Second File to Third

$6$: Dictionary List

$7$: Pandas Basics

$8$: Stats & Pandas

$9$: Data Visualization

$10$: Address Class

Please Enter Homework Number $($and $0$ to exit$)$: $\_$

$2$: Train Dataset $250$ pts

You have to devise an Object Oriented solution for this problem, i$.$e$.$ you MUST create a CLASS, and

solve the problem calling the class methods.

The train dataset contains data regarding a train crash. It has ~$891$ records. However, all records are not

complete, i$.$e$.$ some have missing data. In this data set, our outcome of interest is the survived column,

on whether an individual survived $(1)$ or died $(0)$ after the Train crash.

$1.$ Clean the data and remove any unwanted records. How many records do you have now?

$(20$ points$)$

$2.$ The train picked most passengers from which station $?$

$(20$ points$)$

$3.$ Do some basic data exploration $($e$.$g$.$ using commands as head$(),$ info$(),$ describe$(),$ nunique$(),$ etc$).$

Which variables will you NOT select?

$(20$ points$)$

$4.$ Are there any outliers in the data? If yes, treat them.

$(15$ points$)$

$5.$ Partition the data into a training set $($with $70\%$ of the observations$),$ and testing set $($with $30\%$ of the

observations$)$ using the random state of $12345$ for cross validation. $(15$ points$)$

$6.$ On the partitioned data, build the best KNN model. Show the accuracy numbers. $($Hint: What is the

best value of k$?$ How do you decide the $$best k$?)$

$(30$ points$)$

$7.$ On the partitioned data, build the best logistic regression model. Show the accuracy numbers.

$(30$ points$)$

$8.$ On the partitioned data, build the decision tree. Show the accuracy numbers. What tree depth did

you choose, i$.$e$.$ which one is ideal and why? $(30$ points$)$

$9.$ Based on the results of k$-$nearest neighbor, and logistic regression, what is the best model to classify

the data? Provide explanation to support your argument. $(20$ points$)$

$10.$ Show some interesting graphs of the data, i$.$e$.,$ that can describe the original data.

$(50$ points$)$

Requirements

$$ Make all the "best programming practice" decisions, e$.$g$.$ how to show the output, what prompts

to display, how to ask for input etc.

$$ You are the developer$/$engineer$,$ it is YOUR decision $...$ YOUR job. If the program is not presented

nicely, you will lose points.

$$ First, all that is being asked should be done. Second, the displays should be intuitive, selfexplanatory and nicely put.

Do NOT assume that the user knows ANYTHING.

$$ Write the program such that any new person that sits in front of the terminal, can start playing

with it $($i$.$e$.,$ the commands, displays etc. are adequate$).$

Submission and Demo

MULTIPLE file submissions are allowed. Anything you submit is kept.

You will submit the following:

$.$ipynb files for all questions, clearly labeled. $($Q$1\_$HW$.$ipynb and Q$2\_$Train.ipynb$)$

$$ You are asked to demo the project IN$-$class or Zoom. Details in D$2$L$/$Email$.$

$$ The Demo files $($i$.$e$.$ files you submit$)$ should NOT be commented. python code for this with the detail