机器学习任务汇总.zip

# -*- coding: utf-8 -*- ###Task 1 完成评估函数的定义 ##输入变量为预测值及真实值，返回值相应评估指标，不允许使用除数据结构外的第三方库 # # 真正(True Positive , TP)：被模型预测为正的正样本。--有病，查出来了 # 假正(False Positive , FP)：被模型预测为正的负样本。--没病，说你有病 # 假负(False Negative , FN)：被模型预测为负的正样本。--有病，没查出来（*) # 真负(True Negative , TN)：被模型预测为负的负样本。--没病，检查正常 # # 分类模型的评估: # 准确率Accuracy = (TP+TN)/(TP+TP+FP+FN) # 精确率Precision = (TP)/(TP+FP) # 召回率Recall = (TP)/(TP+FN) # F1-score = (2*Precision*Recall)/(Precision+Recall) predicted_value = [1,0,1,1,0,1,0,1,0,1,1,1,1,1] #预测值 true_value = [1,1,0,1,1,1,0,1,0,1,1,1,1,1] #真实值 1为正样本，0为负样本 def getTruePositive(predicted_value,true_value): tp = 0 for i in range(len(predicted_value)): if true_value[i] == 1 and predicted_value[i] ==1: #被模型预测为正的正样本。 tp += 1 return tp def getFalsePositive(predicted_value,true_value): fp = 0 for i in range(len(predicted_value)): if true_value[i] == 0 and predicted_value[i] == 1: #被模型预测为正的负样本 fp += 1 return fp def getFalseNegative(predicted_value,true_value): fn = 0 for i in range(len(predicted_value)): if true_value[i] == 1 and predicted_value[i] == 0: #被模型预测为负的正样本。 fn += 1 return fn def getTrueNegative(predicted_value,true_value): tn = 0 for i in range(len(predicted_value)): if true_value[i] == 0 and predicted_value[i] == 0: #被模型预测为负的负样本 tn += 1 return tn #precision def metrics_precision(predicted_value,true_value): # 精确率Precision = (TP)/(TP+FP) tp = getTruePositive(predicted_value,true_value) fp = getFalsePositive(predicted_value,true_value) score = tp / (tp+fp) return score #recall def metrics_recall(predicted_value,true_value): # 召回率Recall = (TP) / (TP + FN) tp = getTruePositive(predicted_value,true_value) fn = getFalseNegative(predicted_value,true_value) score = tp/(tp+fn) return score #f1_score def distance_f1_score(predicted_value,true_value): # F1 - score = (2 * Precision * Recall) / (Precision + Recall) precision = metrics_precision(predicted_value, true_value) recall = metrics_recall(predicted_value,true_value) score = 2*precision*recall / (precision+recall) return score #测试precision print(metrics_precision(predicted_value,true_value)) #测试recall print(metrics_recall(predicted_value,true_value)) #测试f1_score print(distance_f1_score(predicted_value,true_value))