day_5_dhruvdhayal_ai_ml.py

# -*- coding: utf-8 -*-
"""Day-5_DHRUVDHAYAL_AI/ML.ipynb

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/19OXvgcpMRY0mhkAEBHN6MeOPG2EvWzOa

#MACHINE LEARNING.
We, will train and test the model such that , model can learn from previous past experiences, surrounding of it's enviornment, by gathering and collecting all data & information in the form of a Knowledge then, it will practically Implement in rel-world implementations to perform all the task and behave just as like a Humans called as "Machine Learning".

#1. Supervised Learning
#2. Unsupervised Learning.
#3. Other Tools and Techniques like: SVM, RainForest Methods, etc.
"""

from sklearn import datasets; #Used for Importing the Valueable Datasets Values.
from sklearn import svm;  #Used for Classification Modelling svm-support vector Machines.
from sklearn import metrics; #Used for Evaluation of the Model.
from sklearn import model_selection; #Used for Model Selection.

#Now, load the main_datasets;
data=datasets.load_digits();
print(type(data));
#print("\n Given Data is: \n\n",data);

#Now, printing the Values of the Data.
print("\n --> Given Values of the Data are: \n\n",data);

#Now, we need to describe the values of the data.
print(data.DESCR);

#Now, after Implementing the Main datasets.
#We, need to divide it ir further in training and testing phase.
#Now, in this, we have to get feature extraction of attributes(data) & labels.
X=data.data;
print(type(X));
print(X.shape);
Y=data.target;
print(type(Y));
print(Y.shape);
#Here, X->Represents the values of the Data and Information in the form of the Attributes.
#Next, Y->Represents the Values of the Labels for each Datasets Attributes.

#Now, we are representing the Values of the datasets data in the form of the Digits.
import matplotlib.pyplot as plt;
import matplotlib.image as mpimg;
import numpy as np;

#Visualise the Data Images.
image=data.images;
y=data.target;
print(type(image));
print(image.shape);
num=1270;
sample_im=image[num,:,:];
label=y[num];
plt.figure(1,figsize=(5,5));
plt.imshow(sample_im,cmap="gray");
plt.title(" Actual Target: "+str(label));
plt.axis("off");

#Now, we are just representing the Values of the Data and Information in the form of the Images present in the given Datasets.
import matplotlib.pyplot as plt;
import matplotlib.image as mpimg;
import numpy as np;

#Visualise the Given Datasets.
image=data.images;
y=data.target;
print(type(image));
print(image.shape);
num=1270;
sample_im=image[num,: ,:];
label=y[num];
plt.figure(1,figsize=(5,5));
plt.imshow(sample_im,cmap="gray");
plt.title(" Acual Target : "+str(label));
plt.axis("off");
plt.show();

"""#Task-1
--> Solve, the Problem Statement that you need to show the collection of data and infromation in datasets in the form of random numbers and represent it in the form of the 7*7 Matrix randomly form of Representations.
"""

#First, we need to implement the values of the Libraries.
import matplotlib.pyplot as plt;
import matplotlib.image as imag;
import numpy as np;

#Visualising the Values of the Data.
images=data.images;
y=data.target;
print("\n Type of the Images in Datasets: ",type(image));
print("\n Total Length of the Images Data are: ",image.shape);

#Now, making the 7*7 Matrix to generate the Random Numbers from the datasets and arrange it in the proper manner.
for i in range(49):
  num=np.random.randint(0,1797);
  samples_im=images[num,: ,: ];
  label=y[num];
  plt.subplot(7,7,i+1);
  plt.figure(1,figsize=(8,6));
  plt.imshow(samples_im,cmap="gray");
  #plt.title(" Actual Target: "+str(label));
  plt.axis("off");

#Now, we need to show the Values of datasets in the form of random Images.
#Importing all the Libraries of it!
import matplotlib.pyplot as plt;
import matplotlib.image as imag;
import numpy as np;

#Visualise the Values of the Data.
image=data.images;
y-data.target;
print("\n --> Type of the Image is: ",type(image));
print("\n --> Total Length of the Images Data are: ",image.shape);
num=1270;
sample_im=image[num, : , :];
label=y[num];
plt.figure(1,figsize=(5,5));
plt.imshow(sample_im,cmap="gray");
plt.title(" Actual Target: \n"+str(label));
plt.axis("off");
plt.show();

#Now, representing all the data and information in the datasets randomly representation by using the (7*7) Matrix.
import matplotlib.pyplot as plt;
import matplotlib.image as imag;
import numpy as np;

#Visualising the values of the Data.
image=data.images;
y=data.target;
print("\n --> Type of the Images present in the Datasets are: ",type(image));
print("\n --> Total Length of the Datasets present in Images are: ",image.shape);

for i in range(49):
  num=np.random.randint(0,1797);
  samples_imp=image[num, : , :];
  label=y[num];
  plt.subplot(7,7,i+1);
  plt.figure(1,figsize=(8,6));
  #plt.title(" Actual Target: "+str(label));
  plt.imshow(samples_imp,cmap="gray");
  plt.axis("off");

#Showing the Values of the main Iamges (7*7) Matrix.
plt.show();

"""#Task-2.
--> Our Main, Task is to generate all the random numbers but in the ascending order.
"""

#Importing all the Required and needed Libraries.
import matplotlib.pyplot as plt  # Fixed import for subplot
import matplotlib.image as imag
import numpy as np;

#Visuaising the Main data values.
image=data.images;
y=data.target;
print("\n --> Type of the Data contained in particular datasets: ",type(image));
print("\n --> Total Size of the Image Datasets are: ",image.shape);

for i in range(100):
  num=np.random.randint(0,1797);
  samples_imp=image[num, : , :];
  label=y[num];
  plt.subplot(10,10,i+1); # Now using plt for subplot
  plt.figure(1,figsize=(9,7));
  plt.imshow(samples_imp,cmap="gray");
  #plt.title(" Actual Target: "+str(label));
  plt.axis("off");

#Showing the Valued Graph given Below!
plt.show();

#Importing all the Required and needed Libraries.
import matplotlib.pyplot as plt  # Fixed import for subplot
import matplotlib.image as imag
import numpy as np;

#Visuaising the Main data values.
image=data.images;
y=data.target;
print("\n --> Type of the Data contained in particular datasets: ",type(image));
print("\n --> Total Size of the Image Datasets are: ",image.shape);

# Figure out the index of each digit
digit_indices = [np.where(y == i)[0] for i in range(10)]

# Plot digits in increasing order
fig, axes = plt.subplots(10, 10, figsize=(9, 7))
for i in range(10):
    for j in range(10):
        index = digit_indices[i][j]
        axes[i, j].imshow(image[index], cmap="gray")
        axes[i, j].axis("off")

#Showing the Graph Finally!
plt.show()

#Now, we have to split the values of the Benchmark.
# like: 60:40 Ratio's.
# Let's we have the 70:40 Ratio, so it contain 70-Testing, 40-Training.
# 80 : 20.
ratio=0.3; #Ratio = 0.3 defines that 30% Data split it into the Testing Phase.
Xtrain,Xtest,ytrain,ytest=model_selection.train_test_split(X,y,test_size=ratio,random_state=5);
print("\n 1. Training Datasets: ",Xtrain.shape," ",ytrain.shape);
print("\n 2. Testing Datasets: ",Xtest.shape,"   ",ytest.shape);

"""#Creating the Classification Model by using the SVM.
#(SVM) Stands for "Support Vector Machine".
"""

#Creating the Classification Model by using the SVM.
# Support vector Machine (SVM).
clf_model=svm.SVC();

#We, need to train the Model by using the Training Datasets.
clf_model=clf_model.fit(Xtrain,ytrain);

#We, need to pass the Trained model in terms of testing.
#we, also mention labels of testing data.
ypred=clf_model.predict(Xtest);

#Showing the Classification of that model.
print(ypred.shape);
print(ytest.shape);

#Now, we need to measure the overall performance and accuracy in order to measure.

acc=metrics.accuracy_score(ypred,ytest);
conf_matrix=metrics.confusion_matrix(ypred,ytest);
report=metrics.classification_report(ypred,ytest);

#=============== Printing and Displaying all the Values! ===============

print("\n 1. Total Accuracy of the Model are: ",acc);
print("\n-------------------------------------------------------");

print("\n 2. Confusion Matrix given here: \n\n",conf_matrix);
print("\n-------------------------------------------------------");

print("\n 3. Classification overall report: \n\n",report);
print("\n-------------------------------------------------------");

"""#Confusion Matrix of that Particular Datasets."""

#We, are showing the Values of the Confusion Matrix are given Below!
import seaborn as sns;
sns.heatmap(conf_matrix,annot=True,cmap='jet');
plt.title(" Confusion Matrix showing the Digit Datasets!");
plt.show();

"""#Task-4
--> 1. Load the iris Data-Sets.

--> 2. Understand the Dataset and describe by 'DESCR'.

--> 3. And Implement the 70:30 & 60:40 Train Test Split.

--> 4. And Compare it's Accuracy, Precision, Reacll & F1-Score.

--> 5. For, Each Split.
"""

#Now, we are the the Iris-Datasets.
from sklearn import datasets;
from sklearn import svm;
from sklearn import model_selection;
from sklearn import metrics;

#Visualise and load the Datasets Simply!
data=datasets.load_iris();
print(type(data));

#Now, we have to show the Values of the Data.
print("\n 1. Given Iris Datasets after Loading properly Given Below! : \n\n",data);
#print("\n 2. Total Length of the Iris-Datasets are: ",data.shape);

#Understand the Dataset and describe by 'DESCR'.
#Describing the Datasets are by using the DESCR.
print(data.DESCR);

#Now, we splitting the Datasets into ration of Testing and Training Phases.
ratio=0.3;
Xtest,Xtrain,ytest,ytrain=model_selection.train_test_split(X,y,test_size=ratio,random_state=5);
print("\n 1. Training Datasets: ",Xtrain.shape," ",ytrain.shape);
print("\n 2. Testing Datasets: ",Xtest.shape,"   ",ytest.shape);

#Now, we need to train the Modelby using the SVM.
#SUPORT VECTOR MACHINE - (SVM) generally used for classification of the Models.
clf_model=svm.SVC();

#Train the Model in terms of training.
clf_model=clf_model.fit(Xtrain,ytrain);

#Now, mention the labels in testing for classifications.
#Now, we used to predict by the trained model.
ypred=clf_model.predict(Xtest);

#Showing the Actual Label and predicted Labels.
print("\n Predicted Values: ",ypred.shape);
print("\n Testing Values: ",ytest.shape);

#Now, we need to compare it's valius of the Accuracy, Precision, and it's F1-Score.

acc=metrics.accuracy_score(ypred,ytest);
confu_matrix=metrics.confusion_matrix(ypred,ytest);
report=metrics.classification_report(ypred,ytest);

#Printing and Displaying all the Values in the Iris Classifications are given Below!

print("\n---------------------- IRIS DATASETS PERFORMANCE ----------------------");

print("\n 1. Total Accuracy are: ",acc);
print("\n---------------------------------------------------------");

print("\n 2. Confusion Matrix are Given Below! \n\n",confu_matrix);
print("\n---------------------------------------------------------");

print("\n 3. Classification Report have been given Below! \n\n",report);
print("\n---------------------------------------------------------");

# We, also show the values of the Confusion Matrix.
import seaborn as sns;
import matplotlib.pyplot as plt;

sns.heatmap(confu_matrix,annot=True,cmap="jet");
plt.title("Iris Recognization Datasets with Labels!");

#Showing the finall plotting of the Graph.
plt.figure(1,figsize=(10,8));
plt.show();

# load the iris dataset
# understand the datset from its DESCR
# and implement the 60:40 and 70:30 ratio train test split
# and compare its acc, precision, recall and f1_score
# for each split

from sklearn import metrics,datasets,model_selection,svm
data = datasets.load_iris()

X = data.data
y = data.target
Result = np.zeros((2,4))
ratio= [0.4,0.3]
for i in range(len(ratio)):
  Xtrain,Xtest,ytrain,ytest = model_selection.train_test_split(X,y,test_size=ratio[i],random_state=5)
  # create the model
  clf_model = svm.SVC()
  # train the model
  clf_model = clf_model.fit(Xtrain,ytrain)
  # test the model
  ypred = clf_model.predict(Xtest)

  # accuracy
  temp_acc = metrics.accuracy_score(ypred,ytest)
  # precision
  temp_pre = metrics.precision_score(ypred,ytest,average='macro')
  # recall
  temp_rec = metrics.recall_score(ypred,ytest,average='macro')
  # f1 score
  temp_f1 = metrics.f1_score(ypred,ytest,average='macro')

  list_result = [temp_acc,temp_pre,temp_rec,temp_f1]
  Result[i,:]=list_result

print(Result)
res = ['Accuracy','Precision','Recall','F1_score']

import pandas as pd
res_df = pd.DataFrame(Result,columns=res,index=['60:40','70:30'])
print(res_df)

res_df.T.plot(kind = 'bar')

"""#----Here,isthe END OF DAY-5 Based on SupervisedLearning.----
#========================== DAY-5 ===========================
"""