محاولة لإستكشاف افضل الطرق لتحسين اداء نموذج بيما



In [1]:

    
import numpy as np
import pandas as pd
import seaborn as sb
from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn import preprocessing

%matplotlib inline

df = pd.read_csv('diabetes.csv')
df.head(20) #لاستعراض ال20 السجلات الاولى من إطار البيانات









    Out[1]:






  
    
      
      Pregnancies
      Glucose
      BloodPressure
      SkinThickness
      Insulin
      BMI
      DiabetesPedigreeFunction
      Age
      Outcome
    
  
  
    
      0
      6
      148
      72
      35
      0
      33.6
      0.627
      50
      1
    
    
      1
      1
      85
      66
      29
      0
      26.6
      0.351
      31
      0
    
    
      2
      8
      183
      64
      0
      0
      23.3
      0.672
      32
      1
    
    
      3
      1
      89
      66
      23
      94
      28.1
      0.167
      21
      0
    
    
      4
      0
      137
      40
      35
      168
      43.1
      2.288
      33
      1
    
    
      5
      5
      116
      74
      0
      0
      25.6
      0.201
      30
      0
    
    
      6
      3
      78
      50
      32
      88
      31.0
      0.248
      26
      1
    
    
      7
      10
      115
      0
      0
      0
      35.3
      0.134
      29
      0
    
    
      8
      2
      197
      70
      45
      543
      30.5
      0.158
      53
      1
    
    
      9
      8
      125
      96
      0
      0
      0.0
      0.232
      54
      1
    
    
      10
      4
      110
      92
      0
      0
      37.6
      0.191
      30
      0
    
    
      11
      10
      168
      74
      0
      0
      38.0
      0.537
      34
      1
    
    
      12
      10
      139
      80
      0
      0
      27.1
      1.441
      57
      0
    
    
      13
      1
      189
      60
      23
      846
      30.1
      0.398
      59
      1
    
    
      14
      5
      166
      72
      19
      175
      25.8
      0.587
      51
      1
    
    
      15
      7
      100
      0
      0
      0
      30.0
      0.484
      32
      1
    
    
      16
      0
      118
      84
      47
      230
      45.8
      0.551
      31
      1
    
    
      17
      7
      107
      74
      0
      0
      29.6
      0.254
      31
      1
    
    
      18
      1
      103
      30
      38
      83
      43.3
      0.183
      33
      0
    
    
      19
      1
      115
      70
      30
      96
      34.6
      0.529
      32
      1

هذه الدالة تعطينا توصيف كامل للبيانات و تكشف لنا في ما إذا كانت هناك قيم مفقودة



In [3]:

    
df.info()









    



<class 'pandas.core.frame.DataFrame'>
RangeIndex: 768 entries, 0 to 767
Data columns (total 9 columns):
Pregnancies                 768 non-null int64
Glucose                     768 non-null int64
BloodPressure               768 non-null int64
SkinThickness               768 non-null int64
Insulin                     768 non-null int64
BMI                         768 non-null float64
DiabetesPedigreeFunction    768 non-null float64
Age                         768 non-null int64
Outcome                     768 non-null int64
dtypes: float64(2), int64(7)
memory usage: 54.1 KB

سيبورن مكتبة جميلة للرسوميات سهلة في الكتابة لكن مفيدة جداً في المعلومات التي ممكن ان نقراءها عبر الهيستوقرام

فائدها ممكن ان تكون في

1- تلخيص توزيع البينات في رسوميات

2- فهم او الإطلاع على القيم الفريدة

3- تحمل الرسوميات معنى اعمق من الكلمات



In [4]:

    
sb.countplot(x='Outcome',data=df, palette='hls')









    Out[4]:





<matplotlib.axes._subplots.AxesSubplot at 0xb5372e8>



In [5]:

    
sb.countplot(x='Pregnancies',data=df, palette='hls')









    Out[5]:





<matplotlib.axes._subplots.AxesSubplot at 0xb640198>



In [6]:

    
sb.countplot(x='Glucose',data=df, palette='hls')









    Out[6]:





<matplotlib.axes._subplots.AxesSubplot at 0xc416cc0>



In [7]:

    
sb.heatmap(df.corr())









    Out[7]:





<matplotlib.axes._subplots.AxesSubplot at 0xb872cf8>



In [8]:

    
sb.pairplot(df, hue="Outcome")









    Out[8]:





<seaborn.axisgrid.PairGrid at 0xcac5080>



In [12]:

    
from scipy.stats import kendalltau
sb.jointplot(df['Pregnancies'], df['Glucose'], kind="hex", stat_func=kendalltau, color="#4CB391")









    Out[12]:





<seaborn.axisgrid.JointGrid at 0x17c8dba8>



In [31]:

    
import matplotlib.pyplot as plt
g = sb.FacetGrid(df, row="Pregnancies", col="Outcome", margin_titles=True)
bins = np.linspace(0, 50, 13)
g.map(plt.hist, "BMI", color="steelblue", bins=bins, lw=0)









    Out[31]:





<seaborn.axisgrid.FacetGrid at 0x657146d8>



In [33]:

    
sb.pairplot(df, vars=["Pregnancies", "BMI"])









    Out[33]:





<seaborn.axisgrid.PairGrid at 0x2574b1d0>

تجربة استخدام تقييس و تدريج الخواص لتحسين اداء النموذج



In [2]:

    
columns = ['Pregnancies', 'Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'DiabetesPedigreeFunction', 'Age']
labels = df['Outcome'].values
features = df[list(columns)].values

X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.30)

clf = RandomForestClassifier(n_estimators=1)
clf = clf.fit(X_train, y_train)

accuracy = clf.score(X_train, y_train)
print ' اداء النموذج في عينة التدريب بدقة ', accuracy*100

accuracy = clf.score(X_test, y_test)
print ' اداء النموذج في عينة الفحص بدقة ', accuracy*100

ypredict = clf.predict(X_train)
print '\n Training classification report\n', classification_report(y_train, ypredict)
print "\n Confusion matrix of training \n", confusion_matrix(y_train, ypredict)

ypredict = clf.predict(X_test)
print '\n Testing classification report\n', classification_report(y_test, ypredict)
print "\n Confusion matrix of Testing \n", confusion_matrix(y_test, ypredict)









    



 اداء النموذج في عينة التدريب بدقة  89.0130353818
 اداء النموذج في عينة الفحص بدقة  70.5627705628

 Training classification report
             precision    recall  f1-score   support

          0       0.90      0.93      0.92       348
          1       0.87      0.81      0.84       189

avg / total       0.89      0.89      0.89       537


 Confusion matrix of training 
[[324  24]
 [ 35 154]]

 Testing classification report
             precision    recall  f1-score   support

          0       0.74      0.85      0.79       152
          1       0.60      0.43      0.50        79

avg / total       0.69      0.71      0.69       231


 Confusion matrix of Testing 
[[129  23]
 [ 45  34]]

تجربة تحسين اداء النموذج باستخدام طريقة

standard scaler



In [41]:

    
#scaling
scaler = StandardScaler()

# Fit only on training data
scaler.fit(X_train)
X_train = scaler.transform(X_train)
# apply same transformation to test data
X_test = scaler.transform(X_test)



In [3]:

    
clf = RandomForestClassifier(n_estimators=1)
clf = clf.fit(X_train, y_train)

accuracy = clf.score(X_train, y_train)
print ' اداء النموذج في عينة التدريب بدقة ', accuracy*100

accuracy = clf.score(X_test, y_test)
print ' اداء النموذج في عينة الفحص بدقة ', accuracy*100

ypredict = clf.predict(X_train)
print '\n Training classification report\n', classification_report(y_train, ypredict)
print "\n Confusion matrix of training \n", confusion_matrix(y_train, ypredict)

ypredict = clf.predict(X_test)
print '\n Testing classification report\n', classification_report(y_test, ypredict)
print "\n Confusion matrix of Testing \n", confusion_matrix(y_test, ypredict)









    



 اداء النموذج في عينة التدريب بدقة  88.6405959032
 اداء النموذج في عينة الفحص بدقة  73.1601731602

 Training classification report
             precision    recall  f1-score   support

          0       0.92      0.91      0.91       348
          1       0.83      0.85      0.84       189

avg / total       0.89      0.89      0.89       537


 Confusion matrix of training 
[[315  33]
 [ 28 161]]

 Testing classification report
             precision    recall  f1-score   support

          0       0.80      0.80      0.80       152
          1       0.61      0.61      0.61        79

avg / total       0.73      0.73      0.73       231


 Confusion matrix of Testing 
[[121  31]
 [ 31  48]]

تجربة تحسين اداء النموذج بطريقة

min-max scaler



In [4]:

    
columns = ['Pregnancies', 'Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'DiabetesPedigreeFunction', 'Age']
labels = df['Outcome'].values
features = df[list(columns)].values

X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.30)

scaler = preprocessing.MinMaxScaler()

scaler.fit(X_train)
X_train = scaler.transform(X_train)
# apply same transformation to test data
X_test = scaler.transform(X_test)

clf = RandomForestClassifier(n_estimators=1)
clf = clf.fit(X_train, y_train)

accuracy = clf.score(X_train, y_train)
print ' اداء النموذج في عينة التدريب بدقة ', accuracy*100

accuracy = clf.score(X_test, y_test)
print ' اداء النموذج في عينة الفحص بدقة ', accuracy*100

ypredict = clf.predict(X_train)
print '\n Training classification report\n', classification_report(y_train, ypredict)
print "\n Confusion matrix of training \n", confusion_matrix(y_train, ypredict)

ypredict = clf.predict(X_test)
print '\n Testing classification report\n', classification_report(y_test, ypredict)
print "\n Confusion matrix of Testing \n", confusion_matrix(y_test, ypredict)









    



 اداء النموذج في عينة التدريب بدقة  89.5716945996
 اداء النموذج في عينة الفحص بدقة  70.5627705628

 Training classification report
             precision    recall  f1-score   support

          0       0.90      0.94      0.92       350
          1       0.88      0.81      0.84       187

avg / total       0.90      0.90      0.89       537


 Confusion matrix of training 
[[329  21]
 [ 35 152]]

 Testing classification report
             precision    recall  f1-score   support

          0       0.76      0.80      0.78       150
          1       0.59      0.53      0.56        81

avg / total       0.70      0.71      0.70       231


 Confusion matrix of Testing 
[[120  30]
 [ 38  43]]



In [18]:

    
columns = ['Pregnancies', 'Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'DiabetesPedigreeFunction', 'Age']
labels = df['Outcome'].values
features = df[list(columns)].values

X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.30)

clf = RandomForestClassifier(n_estimators=5)
clf = clf.fit(X_train, y_train)

accuracy = clf.score(X_train, y_train)
print ' اداء النموذج في عينة التدريب بدقة ', accuracy*100

accuracy = clf.score(X_test, y_test)
print ' اداء النموذج في عينة الفحص بدقة ', accuracy*100

ypredict = clf.predict(X_train)
print '\n Training classification report\n', classification_report(y_train, ypredict)
print "\n Confusion matrix of training \n", confusion_matrix(y_train, ypredict)

ypredict = clf.predict(X_test)
print '\n Testing classification report\n', classification_report(y_test, ypredict)
print "\n Confusion matrix of Testing \n", confusion_matrix(y_test, ypredict)









    



 اداء النموذج في عينة التدريب بدقة  96.834264432
 اداء النموذج في عينة الفحص بدقة  74.025974026

 Training classification report
             precision    recall  f1-score   support

          0       0.96      0.99      0.98       348
          1       0.98      0.93      0.95       189

avg / total       0.97      0.97      0.97       537


 Confusion matrix of training 
[[344   4]
 [ 13 176]]

 Testing classification report
             precision    recall  f1-score   support

          0       0.81      0.79      0.80       152
          1       0.61      0.65      0.63        79

avg / total       0.74      0.74      0.74       231


 Confusion matrix of Testing 
[[120  32]
 [ 28  51]]



In [ ]:

	Pregnancies	Glucose	BloodPressure	SkinThickness	Insulin	BMI	DiabetesPedigreeFunction	Age	Outcome
0	6	148	72	35	0	33.6	0.627	50	1
1	1	85	66	29	0	26.6	0.351	31	0
2	8	183	64	0	0	23.3	0.672	32	1
3	1	89	66	23	94	28.1	0.167	21	0
4	0	137	40	35	168	43.1	2.288	33	1
5	5	116	74	0	0	25.6	0.201	30	0
6	3	78	50	32	88	31.0	0.248	26	1
7	10	115	0	0	0	35.3	0.134	29	0
8	2	197	70	45	543	30.5	0.158	53	1
9	8	125	96	0	0	0.0	0.232	54	1
10	4	110	92	0	0	37.6	0.191	30	0
11	10	168	74	0	0	38.0	0.537	34	1
12	10	139	80	0	0	27.1	1.441	57	0
13	1	189	60	23	846	30.1	0.398	59	1
14	5	166	72	19	175	25.8	0.587	51	1
15	7	100	0	0	0	30.0	0.484	32	1
16	0	118	84	47	230	45.8	0.551	31	1
17	7	107	74	0	0	29.6	0.254	31	1
18	1	103	30	38	83	43.3	0.183	33	0
19	1	115	70	30	96	34.6	0.529	32	1