The solution which I presented just now works perfectly fine on the live editor of Matlab, but don't know why math works still say it is incorrect?
use fix instead of floor or ceil command
Test  Status  Code Input and Output 

1  Fail 
x = 8.8;
y_correct = 8;
assert(isequal(round_zero(x),y_correct))
x =
8

2  Fail 
x = 8.8;
y_correct = 8;
assert(isequal(round_zero(x),y_correct))
x =
8

3  Fail 
x = 0.8;
y_correct = 0;
assert(isequal(round_zero(x),y_correct))
x =
0

4  Fail 
x = 0.4;
y_correct = 0;
assert(isequal(round_zero(x),y_correct))
x =
0

5  Fail 
x = 0;
y_correct = 0;
assert(isequal(round_zero(x),y_correct))
x =
0

6  Fail 
x = eps;
y_correct = 0;
assert(isequal(round_zero(x),y_correct))
x =
0

7  Fail 
x = pi;
y_correct = 3;
assert(isequal(round_zero(x),y_correct))
x =
3

820 Solvers
Project Euler: Problem 1, Multiples of 3 and 5
1121 Solvers
Find nearest prime number less than input number
198 Solvers
Create a two dimensional zero matrix
281 Solvers
127 Solvers