Good Driver Bad Driver
We are starting a new service company providing on-rent good drivers, by the name BBDrivers. We already have 3600 drivers that we have classified as Good=0, Average=1 or Bad=2 and also have their scores on particular tests. We have received 400 more applications for drivers, but reviewing them will be a loss of time and money. Can you please let me know what kind of drivers these guys are? I’ll even give you the old driver’s data in case you need it.
Test your results here (Follow instructions shown on the website) : https://driver.vishwactf.com/ Enter the results only in a string as mentioned on the website. You will get a flag only if you have an accuracy of 1.0 in predicting the classes.
Designer : http://www.antrixacademy.com/
Ok, we got two
csv files. One with labeled data and second with unlabeled (which needs to be labeled in accurate way).
My first thought was to use
scikit-learn, but after initial examination of provided data I decided to first try the
import csv with open('drivertestunlabeled.csv', newline='') as csvfile: csvreader = csv.reader(csvfile, delimiter=',', quotechar='|') next(csvreader, None) for row in csvreader: if float(row) > 100 and float(row) < 30: print('0', end='') elif float(row) > 100 and float(row) > 30: print('2', end='') else: print('1', end='')
I got below as an output.
11101111110110110100111211011111111111110111011001110011011100111111111010111010111111111000111101001101111111111101111 111111011111111011111111111001110010101111121011111101111101111001111111111111111101001101001111110121101001111111111111 111210111111111011111111111101111121111011111111111111111011111100112101110011110111001211111111111111110111111211111111 01111111111111111110110111110111011102111
And it was enough to get the flag ;-)