Uploaded data of Day(3,4,5) Successfully! 😎

DAY-5_SUMMER_TRAINING_AIML/day_5_dhruvdhayal_ai_ml.py

@@ -0,0 +1,384 @@
+# -*- 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 ===========================
+"""