Python Pandas: Read_excel guesses the datatype, how can I protect the column names(headers) from being changed to floats?

Question

I have a dataset in which column headers are numbers (some are integer, some are with fractions), when I use read_excel, it reads all of them as floats, but I do not want the on