考核成绩分类

《考核成绩分类》由会员分享，可在线阅读，更多相关《考核成绩分类（8页珍藏版）》请在装配图网上搜索。

1、Logistic RegressionThe data我们将建立一个逻辑回归模型来预测一个学生是否被大学录取。假设你是一个大学系的管理 员，你想根据两次考试的结果来决定每个申请人的录取机会。你有以前的申请人的历史数据, 你可以用它作为逻辑回归的训练集。对于每一个培训例子，你有两个考试的申请人的分数和 录取决定。为了做到这一点，我们将建立一个分类模型，根据考试成绩估计入学概率。#三大件import numpy as npimport pandas as pdimport matplotlib.pyplot as plt%matplotlib inlineimport os#path = data

2、 + os.sep + LogiReg_data.txtpdData = pd.read_csv(F:研究课程机器学习宇迪系列Wpython数据分析与机器学习实战- 全-E唐宇迪-机器学习课程资料机器学习算法配套案例实战梯度下降 WdataWLogiReg_data.txt, header=None, names= Exam 1, Exam 2, Admitted)pdData.head()Ejcam 1Exam 2Admitted034.623G607&.0246930130.28671143.3943930235 84740972.90219803fflM 32599&6.30855214

3、7.03273&75.3413761pdData.shape(100r 2)positive = pdDatapdData Admitted = 1 # returns the subset of rows such Admitted = 1, i.e. the set of *positive* examplesnegative = pdDatapdData Admitted = 0 # returns the subset of rows such Admitted = 0, i.e. the set of *negative* examplesfig, ax = plt.subplots

4、(figsize=(10,5)ax.scatter(positive Exam 1, positive Exam 2, s=30, c=b, marker=o, label=Admitted) ax.scatter(negative Exam 1, negative Exam 2, s=30, c=r, marker=x, label=Not Admitted) ax.legend()ax.set_xlabel(Exam 1 Score)ax.set_ylabel(Exam 2 Score)rflrtplotlibB1;errt, T thresholdreturn abs(value-1-v

5、alue-2) threshold return np.linalg.norm(value) = n:k = 0X, y = shuffleData(data) #重新洗牌theta = theta - alpha*grad # 参数更新costs.append(cost(X, y, theta) # 计算新的损失i += 1if stopType = STOP_ITER:value = ielif stopType = STOP_COST:value = costselif stopType = STOP_GRAD:value = gradif stopCriterion(stopType,

6、 value, thresh): breakreturn theta, i-1, costs, grad, time.time() - init_timedef runExpe(data, theta, batchSize, stopType, thresh, alpha):#import pdb; pdb.set_trace();theta, iter, costs, grad, dur = descent(data, theta, batchSize, stopType, thresh, alpha) name = Original if (data:,12).sum() 1 else S

7、caledname += data - learning rate: - .format(alpha)if batchSize=n: strDescType = Gradientelif batchSize=1: strDescType = Stochasticelse: strDescType = Mini-batch ().format(batchSize)name += strDescType + descent - Stop: if stopType = STOP_ITER: strStop = iterations.format(thresh)elif stopType = STOP

8、_COST: strStop = costs change .format(thresh)else: strStop = gradient norm .format(thresh)name += strStopprint (*nTheta: - Iter: - Last cost: :03.2f - Duration: :03.2fs.format( name, theta, iter, costs-1, dur)fig, ax = plt.subplots(figsize=(12,4)ax.plot(np.arange(len(costs), costs, r)ax.set_xlabel(I

9、terations)ax.set_ylabel(Cost)ax.set_title(name.upper() + - Error vs. Iteration)return theta不同的停止策略设定迭代次数#选择的梯度下降方法是基于所有样本的n=100runExpe(orig_data, theta, n, STOP_ITER, thresh=5000, alpha=0.000001)data. = leamon： iatr. le-Ofi - G-xadicnt drecent - Slap: ECCO Lt 匕匚乩tiinurO-CDDZTn? H OJTOK32 0.00376111

10、- Itei. 5000 - Lat- . 0. S3 - ImratiOD. L LB=ajrrayftC-ClKOaTia?! Cl OJTDEg島 0. OD37bT L JORIGINAL DATA - LEARNING RATE:： IE-06 亠 GRADIENT DESCENT - STOP: 5000 ITERATIONS Error vs. Iteration根据损失值停止设定阈值1E-6,差不多需要110 000次迭代runExpe(orig_data, theta, n, STOP_COST, thresh=0.000001, alpha=0.001)*f*Orisina

11、l data - leaminc rate: 0, OCd - Gradient dMcwit _ Stop- owts chan st L ill clll k： ：7-.-JJI U. .-Ibeta. -1.37：E3D? 102721632 0.018990) - Iter: O45 - L an cost O.J； - ZuiuLi.u. It.T-yi-：in-i. n. -=. r ：)ORIGIMAL DATA LEARNING KATE: 0.001 - GRADIENT DESCENT . STOP: GRADIENT MOftM = 0.5 else 0 for x in model(X, theta)scaled_X = scaled_data:, :3y = scaled_data:, 3predictions = predict(scaled_X, theta)correct = 1 if (a = 1 and b = 1) or (a = 0 and b = 0) else 0 for (a, b) in zip(predictions, y) accuracy = (sum(map(int, correct) % len(correct)print (accuracy = 0%.format(accuracy)accuracy = 89S