def main():
real_data = neigh.read_trainer('wdbc.data')
synth_data = neigh.make_synth()
n = 10
classifier = neigh.NNclassifier
real_perf = neigh.n_validator(real_data, n, classifier)
synth_perf = neigh.n_validator(synth_data, n, classifier)
print()
print('The model\'s ' + str(n) + '-fold validated performance with real data is ' + str(real_perf)[:5] + '.')
print('The model\'s ' + str(n) + '-fold validated performance with synthetic data is ' + str(synth_perf) + '.')
return
def main():
real_data = neigh.read_trainer('wdbc.data')
synth_data = neigh.make_synth()
n = 10
classifier = neigh.KNNclassifier
norm = [neigh.np.inf, 1] # Maximum norm and l-1 norm
best_real_k = 1
best_real_perf = 0
best_synth_k = 1
best_synth_perf = 0
for m in norm:
for k in range(1, 16, 2):
real_perf = neigh.n_validator(real_data, n, classifier, k, m)
if real_perf > best_real_perf:
best_real_perf = real_perf
best_real_k = k
synth_perf = neigh.n_validator(synth_data, n, classifier, k, m)
if synth_perf > best_synth_perf:
best_synth_perf = synth_perf
best_synth_k = k
print()
print('The '
+ str(k) + '-Nearest Neighbor\'s '
+ str(n) + '-fold validated performance with real data using the '
+ str(m) + '-norm is '
+ str(real_perf)[:5] + '.')
print('The '
+ str(k) + '-Nearest Neighbor\'s '
+ str(n) + '-fold validated performance with synthetic data using the '
+ str(m) + '-norm is '
+ str(synth_perf) + '.')
print()
print('The best k for real data with the '
+ str(m) + '-norm was '
+ str(best_real_k) + ' with a performance of '
+ str(best_real_perf)[:5] + '.')
print('The best k for synthetic data with the '
+ str(m) + '-norm was '
+ str(best_synth_k) + ' with a performance of '
+ str(best_synth_perf) + '.')
return