Q$16$: Word$2$vec is one of early neural models for word representation. There are $2$ approaches in

word$2$vec: skip$-$gram and continuous$-$bag$-$of$-$words $($CBOW$).$ Compare and contrast the Skip$-$gram

and Continuous Bag of Words $($CBOW$)$ architectures in Word$2$Vec. Discuss their differences in

terms of training objectives, the context window, and the types of tasks they are suitable for.