this is the entire question

1. A common way to get rid of having to deal with the threshold separately on a perceptron is to add a new feature. This feature always has value one, and you learn a weight for it. Thus, if you have a 100 dimensional problem with a threshold, we solve it as a 101 dimensional problem with threshold at zero. Draw a picture for one dimensional data and a linear separator with a (non-zero) threshold and draw the corresponding picture for the same data, "lifted into two dimensions, with the corresponding linear separator with threshold zero. (Please make sure that the two separators are actually equivalent!) 1. A common way to get rid of having to deal with the threshold separately on a perceptron is to add a new feature. This feature always has value one, and you learn a weight for it. Thus, if you have a 100 dimensional problem with a threshold, we solve it as a 101 dimensional problem with threshold at zero. Draw a picture for one dimensional data and a linear separator with a (non-zero) threshold and draw the corresponding picture for the same data, "lifted into two dimensions, with the corresponding linear separator with threshold zero. (Please make sure that the two separators are actually equivalent!)