Create a randon one dimensional dataset

x $=$ np$.$arange$(0,100)$

y $=$ x $+$ np$.$random.rand$(100)*30$

plt$.$scatter$($x$,$y$)$

Problem $1$

a$)$ Fit a regression line on the dataset created above using sklearn.linear$\_$model LinearRegression, use default hyperparameters

b$)$ Print out the regression intercept and coefficients

c$)$ Plot the data in blue and the regression line in red

d$)$ Predict the y value for x$=45$ and show it on the plot in green.

Problem $2$

a$)$ Fit a regression line on the dataset created above using tensorflow and keras, use default hyperparameters, sequential model, one dense neuron, activation$=$'linear'. Use 'rmsprop' optimizer and mean squared error for the loss function. Use $1000$ epochs.

b$)$ Print out the regression intercept and coefficients

c$)$ Plot the data in blue and the regression line in red

d$)$ Predict the y value for x$=45$ and show it on the plot in green

e$)$ Print the coefficients in the keras model and compare to the scikit learn modelen