Merge branch 'TAHIR0110:main' into maternal_health_risk

TAHIR0110 · Jul 24, 2024 · 5af3716 · 5af3716
2 parents 528b911 + cead717
commit 5af3716
Show file tree

Hide file tree

Showing 20 changed files with 2,399 additions and 0 deletions.
diff --git a/Heartbeat Classification using ECG/Dataset/README.md b/Heartbeat Classification using ECG/Dataset/README.md
@@ -0,0 +1 @@
+### Dataset Link : https://www.kaggle.com/datasets/shayanfazeli/heartbeat
diff --git a/Heartbeat Classification using ECG/Images/00.png b/Heartbeat Classification using ECG/Images/00.png
diff --git a/Heartbeat Classification using ECG/Images/01.png b/Heartbeat Classification using ECG/Images/01.png
diff --git a/Heartbeat Classification using ECG/Images/02.png b/Heartbeat Classification using ECG/Images/02.png
diff --git a/Heartbeat Classification using ECG/Images/03.png b/Heartbeat Classification using ECG/Images/03.png
diff --git a/Heartbeat Classification using ECG/Images/04.png b/Heartbeat Classification using ECG/Images/04.png
diff --git a/Heartbeat Classification using ECG/Images/05.png b/Heartbeat Classification using ECG/Images/05.png
diff --git a/Heartbeat Classification using ECG/Images/06.png b/Heartbeat Classification using ECG/Images/06.png
diff --git a/Heartbeat Classification using ECG/Images/07.png b/Heartbeat Classification using ECG/Images/07.png
diff --git a/Heartbeat Classification using ECG/Images/08.png b/Heartbeat Classification using ECG/Images/08.png
diff --git a/Heartbeat Classification using ECG/Images/09.png b/Heartbeat Classification using ECG/Images/09.png
diff --git a/Heartbeat Classification using ECG/Images/10.png b/Heartbeat Classification using ECG/Images/10.png
diff --git a/Heartbeat Classification using ECG/Images/11.png b/Heartbeat Classification using ECG/Images/11.png
diff --git a/Heartbeat Classification using ECG/Images/12.png b/Heartbeat Classification using ECG/Images/12.png
diff --git a/Heartbeat Classification using ECG/Images/13.png b/Heartbeat Classification using ECG/Images/13.png
diff --git a/Heartbeat Classification using ECG/Images/14.png b/Heartbeat Classification using ECG/Images/14.png
diff --git a/Heartbeat Classification using ECG/Model/README.md b/Heartbeat Classification using ECG/Model/README.md
@@ -0,0 +1,92 @@
+# Heartbeat Classification using ECG
+
+### Goal
+The goal of this project is to create a deep learning model which will predict the heartbeat of the patients from the ECG images.
+
+### Dataset
+The dataset is collected from the following link, https://www.kaggle.com/datasets/shayanfazeli/heartbeat
+
+### What I have done
+Firstly I have visulized the data using different data visualization techniques and after that the model creation is started.
+
+### Libraries used
+1. `Numpy`
+2. `Pandas`
+3. `sklearn`
+4. `Keras`
+5. `Tensorflow`
+6. `matplotlib`
+
+
+# Data Visualization And Conclusion #
+
+### Content of the Training Data set ###
+
+![image](https://user-images.githubusercontent.com/83596240/183436392-f412bd71-0330-4bc5-bcde-41f78583a959.png)
+
+### Different classes/categories along with their frequency ###
+
+![image](https://user-images.githubusercontent.com/83596240/183436723-a8930a81-472e-4cf6-81ee-136cc47db7c8.png)
+
+### Depicting each class in percentage wise using a pie plot ###
+
+![image](https://user-images.githubusercontent.com/83596240/183436966-228c24f1-6e79-4a64-ba47-8aadba9c8c24.png)
+
+We can see that Normal beats as the highest percentage and Fusion beats as the lowest
+
+### Normalizing and Resizing the Training Data set ###
+
+![image](https://user-images.githubusercontent.com/83596240/183437696-c1d54039-6043-4e74-a344-92d872eedadc.png)
+
+The resultant Training Data set that we get is
+
+![image](https://user-images.githubusercontent.com/83596240/183437918-93bcb404-d275-4cff-8deb-12cd9eacb38c.png)
+
+### Updated classes/categories along with their frequency ###
+
+![image](https://user-images.githubusercontent.com/83596240/183438062-ac62948e-fd33-4662-b8f3-4860bdfa9576.png)
+
+As we can see the training dataset is equally balanced with each category having frequency of 16000
+
+### Depicting each class in percentage wise using a pie plot ###
+
+![image](https://user-images.githubusercontent.com/83596240/183438173-e9899fa1-f644-4f1d-852a-2190d20e50d8.png)
+
+### Sample data from each categories ###
+
+![image](https://user-images.githubusercontent.com/83596240/183438386-e32a0d44-788f-4ce1-819d-5666c39df86f.png)
+
+### Checking the Testing Accuracy and Loss ###
+
+![image](https://user-images.githubusercontent.com/83596240/183439120-d751e678-a494-4ac5-ad38-89bd42c65665.png)
+
+The accuracy of the trained model stands at 96.67% and loss is 13.4%
+
+Graphing the training accuracy and loss against the testing acurracy and loss
+
+![image](https://user-images.githubusercontent.com/83596240/183438969-d446f376-582a-4283-a67d-2e7f71ececc6.png)
+
+As the epochs increases the loss decreses and the accuracy increases
+
+### Confusion matrics of the trained model ###
+
+![image](https://user-images.githubusercontent.com/83596240/183439764-f22613f0-68d8-4153-a5ed-85a3db911727.png)
+
+### Training the model with noise to avoid overfitting ###
+
+![image](https://user-images.githubusercontent.com/83596240/183439990-49f9aa6b-b425-432d-b371-9e192fd4a7fa.png)
+
+### Checking the Testing Accuracy and Loss of the new model ###
+
+![image](https://user-images.githubusercontent.com/83596240/183440057-23200f9e-c8c8-49a2-b397-1409695a2d74.png)
+
+The accuracy as reduced to 76.3% and the loss as increased to 132% when we add noise to the dataset
+
+Confusion matrics of the newly trained model
+
+![image](https://user-images.githubusercontent.com/83596240/183440297-ebee69b5-82bd-4ff8-80fe-302b955d9f1a.png)
+
+### This the Confusion metrics that we get after applying transfer learning and performing fine tune to the noisy model ###
+
+![image](https://user-images.githubusercontent.com/83596240/183440719-34265311-faea-46c7-9759-dc38fdf1f345.png)
+
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		### Dataset Link : https://www.kaggle.com/datasets/shayanfazeli/heartbeat