Last Man Standing - Less than 12 Hrs to GO!

machine_learning
hackathon
analytics

#1

Analytics Vidhya is back with its Signature Hackathon and it is going to be our biggest ever!

The best data scientists across the globe have already registered! Now its your turn.

So you either learn or you win!
Prizes worth Rs,30000 (~$450) are up for grabs. Your success may also lead you to a dream job interview.

Register Now – It’s now or never!

If you have any doubts, then we have arranged an Exclusive Platform where you can ask all your doubts directly to Mr. Kunal (Founder of Analytics Vidhya). LIVE AMA TODAY from 8:30 PM to 9:30 PM (IST).

Login to ASK ANYTHING


#2

#3

Hi, Is the team formation mandatory? can anyone participate as individual (1 team member)?


#4

@afzal

They are allowing to participate as individual.


#5

“The public leaderboard is based on 30% of test data, while final rank would be decided on remaining 70% of test data (which is private leaderboard”

Can anyone explain the meaning of the above lines.


#6

I think it just means that the leaderboard ranking i based on this 30% data. One can be top ranked on this 30% data. But final ranking will be determined using the 70% data which we will not see or be able to fine tune our algorithms on.


#8

What is Number_Weeks_Quit? For example if Number_Weeks_Used = 10 and Number_Weeks_Quit = 2 does it mean they used for 10 weeks and didn’t use of 2 weeks during those 10 weeks? But in the data Number_Weeks_Quit can be > Number_Weeks_Used. So i don’t get it. Did anyone figure it out?


#10

Lets say the test file has 100 rows. The prediction accuracy is calculated(from the .csv file that you uplaod) basing on 30 randomly selected records(which will be constant throughout the competition), this gives you a leader-board(LB) score which is called the Public LB score. The Private LB score is what you get after the end of the competition where your solution will be compared with all the 100 records in the test set.

Hope its clear. :slightly_smiling:


#11

I too have the same question…


#12

Hi,

Can someone help me to understand the Num_Weeks_Used, Num_Weeks_Quit and Num_Doses_Week. I am assuming if Num_Doses_Week is zero that means Num_Weeks_Used must be zero, but I found that for some instances Num_Doses_Week is zero but Num_Weeks_Used is positive and Num_Weeks_Quit i zero. Thanks in advance.

Cheers,
Naveen


#13

HI Nagesh,

Thanks for the explanation. So it means that if i am getting a score of 0.9 on leader board in 30% of the test data, in the final ranking which includes the other 70% test data i might score less or more. If this is true, how can i be sure that i am going in the right direction. Why this 30% and 70% split is necessary.


#15

Hi Divye,

This is where the concept of Cross Validation(CV) comes in. CV gives you a score of the model fit. The score is similar to the LB score(in some situations). I only trust my CV scores. Most of the times it pays off well.


#16

@kunal, would like to know more about these variables. Please help. :slightly_smiling:


#17

@afzal,

No, Team formation is not mandatory.

Regards,
Sunil


#18

@rajgot

If Number_Weeks_Used=10 and Number_Week_Quit=2, it means they used pesticides for 10 weeks but for last 2 weeks they did not use.

10 Weeks (Pesticides Used) + 2 Weeks (Pesticides not used) = Total 12 weeks

Regards,
Sunil


#19

not able to submit solutions! :frowning: its says success but my solutions are not being submitted!


#20

Can anyone tell how to join the slack groups?! i dont understand how?!


#21

@jaknap32,

First registered for Last Man standing then click on button “Request Slack Invite” on the Hackathon page ( http://datahack.analyticsvidhya.com/contest/last-man-standing/lb ).

Regards,
Sunil


#22

What if, Num_Doses_Week = 0 and Num_Week_Used = 10 ???


#23

Thanks. Thats looks like the most reasonable possibility.