def main():
Result = []
FeatureList = []
TheData = data.ReadData()
TSNEResult = tsne.GetTsneResult(TheData)
print("please input the type you want")
print("1:K means")
print("2:K means implemented by myself(slow)")
print("3:Hierarchical")
num = -1
while True:
try:
num = int(input())
except:
continue
if num == 1:
RunKmeansTrain(TheData, TSNEResult)
break
elif num == 2:
RunSelfTrain(TheData, TSNEResult)
break
elif num == 3:
RunHierTrain(TheData, TSNEResult)
break
tsne.PlotTsneGroundTruth(TheData, TSNEResult, num)
def PaintCorrelation():
TheData = data.CleanData(data.ReadData()).astype(float)
TheData = TheData[[
"MFCCs_ 4", "MFCCs_ 6", "MFCCs_11", "MFCCs_13", "MFCCs_15", "MFCCs_17",
"MFCCs_19", "MFCCs_20", "MFCCs_22", "Family"
]]
a = TheData.corr()
print(a)
plt.subplots(figsize=(17, 17))
plt.rcParams['axes.unicode_minus'] = False
sns.heatmap(a, annot=True, vmax=1, square=True, cmap="Blues")
plt.show()
def RunSVMTrain():
'''
用SVM库训练6个feature
'''
TheData = data.ReadData()
for i in range(len(AllFeatureChoices)):
FeatureList = AllFeatureChoices[i]
Result = svm.SVMTrain(TheData, FeatureList)
Accuracy = (Result["TP"] + Result["TN"]) \
/ (Result["TP"] + Result["FP"] + Result["TN"] + Result["FN"])
Precision = (Result["TP"]) \
/ (Result["TP"] + Result["FP"])
Recall = (Result["TP"]) \
/ (Result["TP"] + Result["FN"])
F1 = (2 * Precision * Recall) \
/ (Precision + Recall)
print("Accuracy: {:.2f}".format(100 * Accuracy))
print("Precision: {:.2f}".format(100 * Precision))
print("Recall: {:.2f}".format(100 * Recall))
print("F1: {:.2f}".format(100 * F1))
print("*" * 100)
import matplotlib.pyplot as plt import seaborn as sns import data import pandas TheData = data.ReadData() a = TheData.corr() print(a) plt.subplots(figsize=(17, 17)) plt.rcParams['axes.unicode_minus'] = False sns.heatmap(a, annot=True, vmax=1, square=True, cmap="Blues") plt.show()