25 KiB
25 KiB
import torch
import torch.nn.functional as F
from torch import nn
from torch.autograd import Variable
import torchvision.transforms as transforms
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics import accuracy_score
import numpy as np
import pandas as pd
class LogisticRegressionModel(nn.Module):
def __init__(self, input_dim, output_dim):
super(LogisticRegressionModel, self).__init__()
self.linear = nn.Linear(input_dim, output_dim)
self.sigmoid = nn.Sigmoid()
def forward(self, x):
out = self.linear(x)
return self.sigmoid(out)
data_train = pd.read_csv("data_train.csv")
data_test = pd.read_csv("data_test.csv")
data_val = pd.read_csv("data_val.csv")
FEATURES = [ 'age','hypertension','heart_disease','ever_married', 'avg_glucose_level', 'bmi']
x_train = data_train[FEATURES].astype(np.float32)
y_train = data_train['stroke'].astype(np.float32)
x_test = data_test[FEATURES].astype(np.float32)
y_test = data_test['stroke'].astype(np.float32)
fTrain = torch.from_numpy(x_train.values)
tTrain = torch.from_numpy(y_train.values.reshape(2945,1))
fTest= torch.from_numpy(x_test.values)
tTest = torch.from_numpy(y_test.values)
batch_size = 95
n_iters = 1000
num_epochs = int(n_iters / (len(x_train) / batch_size))input_dim = 6
output_dim = 1
model = LogisticRegressionModel(input_dim, output_dim)
learning_rate = 0.001
criterion = torch.nn.BCELoss(reduction='mean')
optimizer = torch.optim.SGD(model.parameters(), lr = learning_rate)print(list(model.parameters())[0].size())
print(list(model.parameters())[1].size())torch.Size([1, 6]) torch.Size([1])
for epoch in range(num_epochs):
print ("Epoch #",epoch)
model.train()
optimizer.zero_grad()
# Forward pass
y_pred = model(fTrain)
# Compute Loss
loss = criterion(y_pred, tTrain)
print(loss.item())
# Backward pass
loss.backward()
optimizer.step()Epoch # 0 4.4554009437561035 Epoch # 1 2.887434244155884 Epoch # 2 1.4808591604232788 Epoch # 3 0.6207292079925537 Epoch # 4 0.4031478762626648 Epoch # 5 0.34721270203590393 Epoch # 6 0.32333147525787354 Epoch # 7 0.3105970621109009 Epoch # 8 0.30295372009277344 Epoch # 9 0.2980167269706726 Epoch # 10 0.29466450214385986 Epoch # 11 0.29230451583862305 Epoch # 12 0.29059702157974243 Epoch # 13 0.2893349230289459 Epoch # 14 0.2883857190608978 Epoch # 15 0.2876618504524231 Epoch # 16 0.2871031165122986 Epoch # 17 0.28666743636131287 Epoch # 18 0.28632479906082153 Epoch # 19 0.2860531508922577 Epoch # 20 0.28583624958992004 Epoch # 21 0.2856619954109192 Epoch # 22 0.285521000623703 Epoch # 23 0.2854064106941223 Epoch # 24 0.2853126525878906 Epoch # 25 0.2852354049682617 Epoch # 26 0.2851715385913849 Epoch # 27 0.28511837124824524 Epoch # 28 0.2850736975669861 Epoch # 29 0.2850360572338104 Epoch # 30 0.28500401973724365 Epoch # 31 0.2849765419960022 X:\Anaconda2020\lib\site-packages\torch\autograd\__init__.py:145: UserWarning: CUDA initialization: The NVIDIA driver on your system is too old (found version 10010). Please update your GPU driver by downloading and installing a new version from the URL: http://www.nvidia.com/Download/index.aspx Alternatively, go to: https://pytorch.org to install a PyTorch version that has been compiled with your version of the CUDA driver. (Triggered internally at ..\c10\cuda\CUDAFunctions.cpp:109.) Variable._execution_engine.run_backward(
y_pred = model(fTest)
print("predicted Y value: ", y_pred.data)predicted Y value: tensor([[0.0468],
[0.0325],
[0.2577],
[0.2059],
[0.1090],
[0.0229],
[0.2290],
[0.0689],
[0.2476],
[0.0453],
[0.0150],
[0.4080],
[0.0424],
[0.0981],
[0.0221],
[0.1546],
[0.1400],
[0.1768],
[0.1684],
[0.0229],
[0.1836],
[0.1200],
[0.0137],
[0.2316],
[0.0185],
[0.0179],
[0.0108],
[0.0175],
[0.0471],
[0.4576],
[0.0210],
[0.0103],
[0.0616],
[0.1850],
[0.4114],
[0.4264],
[0.0405],
[0.0788],
[0.2405],
[0.0340],
[0.4345],
[0.1758],
[0.0385],
[0.0749],
[0.4349],
[0.0357],
[0.0295],
[0.3939],
[0.1147],
[0.3812],
[0.0659],
[0.0675],
[0.0263],
[0.1398],
[0.0959],
[0.0406],
[0.0531],
[0.0500],
[0.4259],
[0.1086],
[0.0611],
[0.0855],
[0.0473],
[0.2826],
[0.1734],
[0.0560],
[0.0466],
[0.0290],
[0.1903],
[0.4515],
[0.0118],
[0.2158],
[0.1293],
[0.2488],
[0.0424],
[0.1809],
[0.0122],
[0.0796],
[0.0901],
[0.0879],
[0.0457],
[0.0091],
[0.0196],
[0.3310],
[0.0978],
[0.0843],
[0.0684],
[0.0340],
[0.0583],
[0.0670],
[0.0133],
[0.1165],
[0.0145],
[0.1581],
[0.1677],
[0.0353],
[0.0745],
[0.0108],
[0.0492],
[0.0611],
[0.2977],
[0.2820],
[0.0219],
[0.0580],
[0.0122],
[0.0726],
[0.3315],
[0.0201],
[0.2460],
[0.0110],
[0.0322],
[0.0180],
[0.0135],
[0.3176],
[0.1390],
[0.0678],
[0.1596],
[0.0128],
[0.0900],
[0.0117],
[0.0224],
[0.0357],
[0.0103],
[0.1728],
[0.0135],
[0.0992],
[0.4371],
[0.4525],
[0.0278],
[0.0617],
[0.2499],
[0.0129],
[0.0424],
[0.0292],
[0.3903],
[0.0108],
[0.0404],
[0.0344],
[0.4109],
[0.3936],
[0.0603],
[0.4396],
[0.1155],
[0.3594],
[0.0305],
[0.0307],
[0.0226],
[0.1284],
[0.0474],
[0.0959],
[0.0135],
[0.0289],
[0.3705],
[0.1538],
[0.4535],
[0.0355],
[0.0169],
[0.1648],
[0.4217],
[0.0951],
[0.0767],
[0.0475],
[0.0452],
[0.1625],
[0.0896],
[0.0114],
[0.0423],
[0.3971],
[0.0173],
[0.0250],
[0.3579],
[0.0131],
[0.0201],
[0.0149],
[0.2615],
[0.1773],
[0.1204],
[0.3556],
[0.2390],
[0.0098],
[0.0190],
[0.3040],
[0.0115],
[0.2033],
[0.1327],
[0.0180],
[0.0610],
[0.2927],
[0.1182],
[0.0115],
[0.4474],
[0.3513],
[0.0451],
[0.4089],
[0.0375],
[0.0127],
[0.0630],
[0.0428],
[0.2085],
[0.0529],
[0.3436],
[0.0678],
[0.0717],
[0.0799],
[0.0967],
[0.1246],
[0.1086],
[0.0387],
[0.1742],
[0.1582],
[0.1374],
[0.4205],
[0.0534],
[0.3051],
[0.1204],
[0.0423],
[0.0324],
[0.0141],
[0.0312],
[0.0261],
[0.2619],
[0.0767],
[0.1742],
[0.0311],
[0.1763],
[0.0326],
[0.0529],
[0.3928],
[0.1209],
[0.0724],
[0.3551],
[0.2067],
[0.0288],
[0.0782],
[0.0661],
[0.0469],
[0.1089],
[0.0985],
[0.1032],
[0.1083],
[0.0546],
[0.0983],
[0.1302],
[0.0153],
[0.2179],
[0.1196],
[0.0275],
[0.4366],
[0.0340],
[0.0286],
[0.1193],
[0.0729],
[0.0553],
[0.0159],
[0.0140],
[0.2195],
[0.3792],
[0.3966],
[0.0424],
[0.0872],
[0.0687],
[0.1941],
[0.0179],
[0.0380],
[0.2445],
[0.1905],
[0.1518],
[0.2370],
[0.0706],
[0.1668],
[0.1265],
[0.2363],
[0.0354],
[0.0263],
[0.0653],
[0.0097],
[0.0152],
[0.0495],
[0.2952],
[0.3581],
[0.0388],
[0.0365],
[0.2808],
[0.0189],
[0.0133],
[0.0692],
[0.0256],
[0.0500],
[0.1452],
[0.3315],
[0.4509],
[0.2079],
[0.0140],
[0.2505],
[0.1044],
[0.0121],
[0.0087],
[0.0453],
[0.0173],
[0.0127],
[0.0176],
[0.3826],
[0.0843],
[0.0885],
[0.0133],
[0.0190],
[0.0278],
[0.0612],
[0.2574],
[0.0404],
[0.0735],
[0.0527],
[0.0300],
[0.1384],
[0.2189],
[0.0301],
[0.2303],
[0.0425],
[0.0294],
[0.0103],
[0.2857],
[0.0620],
[0.0938],
[0.2283],
[0.0237],
[0.1175],
[0.1809],
[0.0725],
[0.0107],
[0.0629],
[0.4485],
[0.0233],
[0.0165],
[0.2240],
[0.1200],
[0.0665],
[0.1153],
[0.2957],
[0.0112],
[0.0328],
[0.2823],
[0.4248],
[0.0308],
[0.0216],
[0.1206],
[0.2086],
[0.0115],
[0.0399],
[0.0246],
[0.3960],
[0.1782],
[0.0591],
[0.0092],
[0.0600],
[0.2938],
[0.1144],
[0.0136],
[0.2075],
[0.0426],
[0.0998],
[0.0407],
[0.2944],
[0.2721],
[0.0734],
[0.2927],
[0.0482],
[0.2740],
[0.0363],
[0.4624],
[0.0558],
[0.1669],
[0.0243],
[0.0109],
[0.1209],
[0.0617],
[0.0634],
[0.0183],
[0.0319],
[0.1135],
[0.0121],
[0.0314],
[0.0137],
[0.0195],
[0.0094],
[0.3304],
[0.0694],
[0.0144],
[0.0278],
[0.1393],
[0.3971],
[0.0939],
[0.0489],
[0.0763],
[0.0394],
[0.2953],
[0.0581],
[0.0404],
[0.0489],
[0.0429],
[0.1940],
[0.0098],
[0.0535],
[0.2953],
[0.2188],
[0.0115],
[0.1468],
[0.0210],
[0.2410],
[0.0685],
[0.1935],
[0.1258],
[0.0146],
[0.0279],
[0.0240],
[0.3981],
[0.2131],
[0.0267],
[0.0184],
[0.2806],
[0.2224],
[0.2687],
[0.0207],
[0.2931],
[0.0707],
[0.0408],
[0.0836],
[0.0799],
[0.1043],
[0.0235],
[0.1093],
[0.0915],
[0.0186],
[0.0885],
[0.0143],
[0.0430],
[0.4653],
[0.1440],
[0.0343],
[0.1683],
[0.0222],
[0.1450],
[0.0497],
[0.2624],
[0.0158],
[0.0157],
[0.3439],
[0.1724],
[0.1858],
[0.4211],
[0.0741],
[0.0708],
[0.0437],
[0.0117],
[0.1091],
[0.0450],
[0.1210],
[0.0864],
[0.1131],
[0.0796],
[0.3588],
[0.1135],
[0.0211],
[0.0152],
[0.0109],
[0.1337],
[0.0341],
[0.1293],
[0.0809],
[0.1133],
[0.0163],
[0.0598],
[0.1512],
[0.0480],
[0.1759],
[0.1126],
[0.1127],
[0.0263],
[0.0215],
[0.4364],
[0.0164],
[0.0447],
[0.0979],
[0.0678],
[0.0624],
[0.0317],
[0.0102],
[0.0188],
[0.0698],
[0.1521],
[0.0642],
[0.3642],
[0.4675],
[0.0807],
[0.0191],
[0.0761],
[0.0902],
[0.2640],
[0.0658],
[0.0115],
[0.0166],
[0.0222],
[0.0138],
[0.0588],
[0.0254],
[0.3206],
[0.0467],
[0.0861],
[0.0161],
[0.0726],
[0.2488],
[0.0441],
[0.0508],
[0.2247],
[0.2302],
[0.1532],
[0.4200],
[0.0451],
[0.3172],
[0.3833],
[0.1273],
[0.0189],
[0.0763],
[0.2182],
[0.0644],
[0.2589],
[0.1022],
[0.0138],
[0.0616],
[0.2501],
[0.0143],
[0.0120],
[0.3911],
[0.2098],
[0.1860],
[0.0638],
[0.0235],
[0.0264],
[0.1376],
[0.3966],
[0.2488],
[0.1522],
[0.0390],
[0.3700],
[0.1037],
[0.0729],
[0.1019],
[0.0281],
[0.0292],
[0.0099],
[0.2580],
[0.0105],
[0.1386],
[0.0141],
[0.0188],
[0.0625],
[0.2696],
[0.0582],
[0.0218],
[0.1327],
[0.0290],
[0.3961],
[0.0202],
[0.0209],
[0.3187],
[0.1900],
[0.0237],
[0.3660],
[0.1311],
[0.0300],
[0.0211],
[0.4100],
[0.0311],
[0.1034],
[0.4346],
[0.1150],
[0.0588],
[0.1075],
[0.0989],
[0.2195],
[0.4500],
[0.0934],
[0.0930],
[0.1336],
[0.1932],
[0.1717],
[0.0731],
[0.1601],
[0.2492],
[0.0096],
[0.0759],
[0.1010],
[0.0592],
[0.4519],
[0.2835],
[0.0693],
[0.1462],
[0.0280],
[0.4231],
[0.0400],
[0.1261],
[0.0129],
[0.1344],
[0.0724],
[0.0362],
[0.0444],
[0.0724],
[0.0266],
[0.0624],
[0.0094],
[0.0557],
[0.1328],
[0.1478],
[0.1098],
[0.0486],
[0.1091],
[0.1119],
[0.1213],
[0.2821],
[0.4471],
[0.0485],
[0.0278],
[0.2685],
[0.0907],
[0.0253],
[0.0618],
[0.0361],
[0.2087],
[0.1609],
[0.2896],
[0.2296],
[0.1015],
[0.2034],
[0.0103],
[0.0483],
[0.0843],
[0.2846],
[0.0126],
[0.0839],
[0.0097],
[0.3067],
[0.0319],
[0.1666],
[0.0134],
[0.3563],
[0.0354],
[0.2181],
[0.0399],
[0.1233],
[0.0332],
[0.1251],
[0.3705],
[0.0720],
[0.0260],
[0.2191],
[0.0998],
[0.2293],
[0.1474],
[0.0092],
[0.0092],
[0.0433],
[0.1093],
[0.0165],
[0.3294],
[0.0136],
[0.0735],
[0.0381],
[0.1373],
[0.1616],
[0.0496],
[0.1992],
[0.0342],
[0.2832],
[0.0306],
[0.0188],
[0.3583],
[0.1543],
[0.0188],
[0.0111],
[0.0964],
[0.0963],
[0.2209],
[0.1034],
[0.1088],
[0.0695],
[0.0308],
[0.0280],
[0.0712],
[0.0474],
[0.1890],
[0.3057],
[0.0896],
[0.2190],
[0.1548],
[0.4623],
[0.4395],
[0.3971],
[0.0328],
[0.0132],
[0.4267],
[0.0234],
[0.0202],
[0.0141],
[0.1102],
[0.0159],
[0.0842],
[0.0629],
[0.1334],
[0.1256],
[0.2835],
[0.3958],
[0.1798],
[0.0108],
[0.0144],
[0.0220],
[0.0982],
[0.1031],
[0.0454],
[0.0292],
[0.1306],
[0.4508],
[0.0465],
[0.3683],
[0.0549],
[0.0102],
[0.0222],
[0.0503],
[0.0245],
[0.1466],
[0.1999],
[0.0237],
[0.1147],
[0.4205],
[0.0654],
[0.1033],
[0.1182],
[0.0737],
[0.0561],
[0.2595],
[0.0116],
[0.1922],
[0.4246],
[0.3039],
[0.1907],
[0.4135],
[0.3967],
[0.2716],
[0.2395],
[0.2179],
[0.2798],
[0.0185],
[0.0197],
[0.0482],
[0.0086],
[0.0088],
[0.1054],
[0.1191],
[0.0319],
[0.1223],
[0.0358],
[0.2026],
[0.0206],
[0.0784],
[0.1204],
[0.3416],
[0.3174],
[0.0210],
[0.0305],
[0.0098],
[0.3320],
[0.0258],
[0.3058],
[0.3179],
[0.1205],
[0.0276],
[0.0308],
[0.0135],
[0.3972],
[0.2718],
[0.4641],
[0.0432],
[0.0744],
[0.3558],
[0.2100],
[0.1327],
[0.1907],
[0.0276],
[0.1947],
[0.0844],
[0.2688],
[0.0658],
[0.0929],
[0.1471],
[0.1222],
[0.0212],
[0.0804],
[0.4366],
[0.2472],
[0.0325],
[0.0178],
[0.3040],
[0.0868],
[0.0683],
[0.0470],
[0.2027],
[0.0262],
[0.1257],
[0.1146],
[0.0586],
[0.1017],
[0.4349],
[0.0286],
[0.0723],
[0.3070],
[0.0135],
[0.0380],
[0.0447],
[0.0161],
[0.0729],
[0.0360],
[0.1328],
[0.0126],
[0.0531],
[0.1831],
[0.4434],
[0.0198],
[0.0878],
[0.0382],
[0.0387],
[0.0438],
[0.0720],
[0.0311],
[0.1378],
[0.1178],
[0.1888],
[0.1199],
[0.2023],
[0.1153],
[0.0523],
[0.0420],
[0.0200],
[0.1671],
[0.1086],
[0.0503],
[0.0441],
[0.0242],
[0.1676],
[0.0120],
[0.4497],
[0.0396],
[0.0922],
[0.2931],
[0.0194],
[0.0267],
[0.1875],
[0.1045],
[0.0839],
[0.1251],
[0.0294],
[0.1090],
[0.0136],
[0.0851],
[0.0360],
[0.0158],
[0.0944],
[0.0110],
[0.0114],
[0.0586],
[0.4468],
[0.0760],
[0.0501],
[0.2267],
[0.0528],
[0.0367],
[0.0803],
[0.1456],
[0.2818],
[0.0266],
[0.1995],
[0.3691],
[0.2341],
[0.2593],
[0.0636],
[0.1788],
[0.0479],
[0.0509],
[0.1104],
[0.0918],
[0.1508],
[0.3680],
[0.3948],
[0.1899],
[0.2569],
[0.0363],
[0.0262],
[0.0936],
[0.0550],
[0.1027],
[0.1444],
[0.0330],
[0.0097],
[0.0761],
[0.2207],
[0.0326],
[0.2501],
[0.0394],
[0.0760],
[0.0381],
[0.0115],
[0.2717],
[0.0423],
[0.0731],
[0.1560],
[0.0826],
[0.0092],
[0.0219],
[0.0751],
[0.1322],
[0.2677],
[0.1361],
[0.4089],
[0.0925],
[0.0266],
[0.1068],
[0.3935],
[0.0987],
[0.0115],
[0.3348],
[0.0551],
[0.0817],
[0.0489],
[0.1392],
[0.0596],
[0.0844],
[0.2388],
[0.0960],
[0.0721],
[0.1400],
[0.4667],
[0.2374],
[0.0349],
[0.0857],
[0.1599],
[0.1922],
[0.0281],
[0.0183],
[0.4507],
[0.0167],
[0.0283],
[0.0402],
[0.2076],
[0.1693],
[0.1446],
[0.3547],
[0.0943],
[0.3730],
[0.1823],
[0.0426],
[0.0149],
[0.0327],
[0.3715],
[0.0474],
[0.1343],
[0.1915],
[0.0690],
[0.0092],
[0.1643],
[0.2189],
[0.3149],
[0.2171],
[0.2178],
[0.0097],
[0.3628],
[0.0163],
[0.0684],
[0.1145],
[0.4074],
[0.0514],
[0.3587],
[0.0905],
[0.0159],
[0.1992],
[0.0109]])
print ("The accuracy is", accuracy_score(tTest, np.argmax(y_pred.detach().numpy(), axis=1)))The accuracy is 0.9480651731160896
torch.save(model, 'stroke.pkl')