As a matter of proper coding to develop the most efficient, predictive algorithm, should a code BEGIN with the OUTPUT that appears MOST OFTEN over the ENTIRE SAMPLE of data, and end with the OUTPUT that appears LEAST OFTEN, or vice versa?
To be more specific, I have 6 COLUMNS (A through E) of data and 3400 ROWS (1-3400). I am asking whether EACH of the 6 data points TODAY (A1 through E1) are GREATER than, LESS than, or EQUAL to the 6 data points for the PREVIOUS DAY (A2 through E2). Depending on the answer to each question, each output will receive a simple numeric code. For instance, if A1>A2, then 1. If A1>B2, then 2 etc. The numeric code will simply repeat itself each day. So, if A2>A3, it will also receive a 1 for its code.
My ultimate aim is to determine the extent to which today's ENTIRE code is predictive of tomorrow's ENTIRE code. Therefore, my question is how should the CODE be SEQUENCED?
2006-08-15
10:38:54
·
2 answers
·
asked by
Anonymous