how to compute and plot mean square error for two vectors?

์กฐํšŒ ์ˆ˜: 4(์ตœ๊ทผ 30์ผ)
DEEPAK Chekuri
DEEPAK Chekuri 2021๋…„ 1์›” 9์ผ
๋Œ“๊ธ€: DEEPAK Chekuri 2021๋…„ 1์›” 13์ผ
i have a dataset to classify, using perceptron learning rule . i've calculated the weight matrix but don't know how to plot MSE .
{๐‘1 = [ 1 1 ],๐‘ก1 = [ 0 0 ]}, {๐‘2 = [ 1 2 ],๐‘ก2 = [ 0 0 ]}, {๐‘3 = [ 2 โˆ’1 ],๐‘ก3 = [ 0 1 ]}, {๐‘4 = [ 2 0 ],๐‘ก4 = [ 0 1 ]}, {๐‘5 = [ โˆ’1 2 ],๐‘ก5 = [ 1 0 ]}, {๐‘6 = [ โˆ’2 1 ],๐‘ก6 = [ 1 0 ]}, {๐‘7 = [ โˆ’1 โˆ’1 ],๐‘ก7 = [ 1 1 ]}, {๐‘8 = [ โˆ’2 โˆ’2 ],๐‘ก8 = [ 1 1 ]}.
This the dataset and w=[-2 0;0 -2],bias =[-1 0]

์ฑ„ํƒ๋œ ๋‹ต๋ณ€

Gaurav Garg
Gaurav Garg 2021๋…„ 1์›” 13์ผ
Hi Deepak,
You can plot MSE/Loss and accuracy for each iteration of your training/testing.
To do this, you can make a network with 'n' number of layers, train your network on it and store the loss returned per iteration in a list. Finally, you can plot this loss on y-axis and number of iterations on x-axis.
For any more information on monitoring metrics, you can look at the documentation here.
  ๋Œ“๊ธ€ ์ˆ˜: 1
DEEPAK Chekuri
DEEPAK Chekuri 2021๋…„ 1์›” 13์ผ
Thankyou Gaurav,
Storing mse in a new list for every iteration worked out for me and i'm instruucted to use single layer.
I've stored MSE of each iteration in new list and plotted the same.

๋Œ“๊ธ€์„ ๋‹ฌ๋ ค๋ฉด ๋กœ๊ทธ์ธํ•˜์‹ญ์‹œ์˜ค.

์ถ”๊ฐ€ ๋‹ต๋ณ€(0๊ฐœ)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by