classification - Supervised Machine Learning algorithms -

i trying classify data using supervised machine learning algorithms.

everything's working fine, curiosity, tried 6 classification algorithms simultaneously on single data set. steps followed follows-

1> train algorithms.
2> predicted result(either 1 or 0) test_data individually, algorithms.
3> if of algos gave 0, considered result data pair 0, result 1.
4> found out overall accuracy.

expected overall accuracy higher individual results(by each algorithm working individually), got average accuracy.(average here means average of accuracies of individual algos).

can please me find reason?

this depends on algorithms picked. many algorithms sensitive different things. instance, k-means, linear svm, , power iteration clustering markedly different results.

you got asked for: averaged votes, without coordinating algorithms in way. got average result.

i doubt weighted averaging much; you're doing there training meta-model. instead, consider data set have. need research modelling algorithms , pick 1 tends work on statistical shape of data set respect desired purpose. since you've given none of background, can't specifics.