A. It is known that a 3023/01/atural law obeys the quadratic relationship, Ax2 where predictor x is a continuous random variable, uniformly distributed between -1 and +1 and y is a real valued response variable. What is the best linear curve of the form y=a*x+b that can be used to model this data? That is, find the optimal values of a and b in terms of A that minimize the loss function! B. Consider a table with a single attribute "wind" and category "rain", where "wind" can take two attribute values high and low, and "rain" has two classes yes and no. There are 10 entries in the table, and it is known that 8 entries in the table have wind-high. It is also known that 8 entries in the table also have rain-yes. What is the highest and lowest possible information gain associated with attribute "wind"? 5+5=10 Marks