need help

For the A attribute of pds.arff, consider two possible splits v1 = 3.5 and v2 = 5.5 for discretizing the attribute. Use entropy-based binning to find the information gain for each split and decide which split is better. [If we want to find the optimal split we would need to evaluate all possible splits. Here we simplified the problem.]

work @relation pds @attribute 'A' real @attribute 'B' real @attribute 'C' real @attribute ' positive} @data %A B 0 0 1 4 5 6 6 6 w NN 6 Class O negative o negative 1 negative 2 positive 2 positive 2 negative 2 negative 2 negative 3 positive 3 positive 3 positive 4 positive 5 positive 5 negative 6 positive 7 7 7 7 7 7 0 00 00 8 9 9 10 work @relation pds @attribute 'A' real @attribute 'B' real @attribute 'C' real @attribute ' positive} @data %A B 0 0 1 4 5 6 6 6 w NN 6 Class O negative o negative 1 negative 2 positive 2 positive 2 negative 2 negative 2 negative 3 positive 3 positive 3 positive 4 positive 5 positive 5 negative 6 positive 7 7 7 7 7 7 0 00 00 8 9 9 10