Given the following list of tweets, how many parameters would a logistic regression classifier have to learn with inputs represented as discussed in this lecture? (*Remember to add in 1 )
[“I am happy because I am learning NLP”, “I hated that movie”, “I love working at DL”]
why we have to add 1??
Hey @rugs,
I guess the question itself tells us “why we have to add a 1”. If you read the question in the end, it says
(Remember to add in 1 as well because of the bias)
Logistic regression has a single weight parameter for each of it’s inputs + 1 additional parameter for bias. You will get to know more about this in the upcoming videos which are focused towards describing logistic regression. I hope this helps.
Cheers,
Elemento