How do I normalize the data set of 3 groups and 50 images (3X50) and for each image9 features have been calculated?

Here is the part view of the data

[Joep Linssen]

How do you want to normalize the data? Along each row, along each column, for the entire set? And, what is your normalization criteria; should the maximum singular feature value not exceed 1, or, should the sum of the 9 features not exceed 1, or, something else?

[Bajdar Nour]

Dear @Joep Linssen thanks for the reply..simply I want to determine max. value among all then divides each feature in each image by this max. value.

Answer 1

[Joep Linssen]

Hi Bajdar,

Based on your comment the following code should work:

 feature_vel = cell(3, 50);
 feature_vel = cellfun(@(x) rand(1,9).*2, feature_vel, 'UniformOutput', false);

 max_per_entry = cellfun(@max, feature_vel);
 max_of_all_entries = max(max(max_per_entry));

 feature_vel_normalized = cellfun(@(x) x./max_of_all_entries, feature_vel, 'UniformOutput', false);

The utilized function cellfun applies a function to each element of a cell array. First it is used to compute the maximum value of the 9 features of each element. The max(max()) call then finds the maximum value across all elements. Finally, cellfun is used again to divide the arrays at each element of the cell vector by the found maximum value, by utilizing an anonymous function.

[Bajdar Nour]

Thanks Dear it works