Wiki Report for ICP9 - NagaSurendraBethapudi/Python-ICP GitHub Wiki

Video Link : https://drive.google.com/file/d/1IR8flVYrmYEH0ABegMIuMZgLIAppbw0K/view?usp=sharing

---

Question 1:

Plot the loss and accuracy for both training data and validation data using the history object in the source code (https://umkc.box.com/s/10nrlk6216fncengv7qxbbw5o9vgc3hs)

Answer :

1. Imported the libraries.
2. Imported test and train data from MNIST.
3. Processed the data and built the model.

4. Evaluated result on test data.

5. Loss and Accuracy data plot

---

Question 2:

Plot one of the images in the test data, and then do inferencing to check what is the prediction of the model on that single image (https://umkc.box.com/s/10nrlk6216fncengv7qxbbw5o9vgc3hs

Answer :

1. Imported the libraries.
2. Imported test and train data from MNIST.
3. Processed the data and built the model.

4. Evaluated result on test data.

5. Predicted the unknown data

---

Question 3 :

We had used 2 hidden layers and Relu activation. Try to change the number of hidden layer and the activation to tanh or sigmoid and see what happens. (https://umkc.box.com/s/10nrlk6216fncengv7qxbbw5o9vgc3hs)

Answer :

1. Imported the libraries.
2. Imported test and train data from MNIST.
3. Processed the data and built the model.
4. Modified the model.

5. Evaluated result on test data

6. Loss and Accuracy data plot

7. Conclusion after adding more hidden layers:

• Loss was reduced to 10% from 12%
• Accuracy was same

---

Question 4 :

Run the same code without scaling the images and check the performance (https://umkc.box.com/s/10nrlk6216fncengv7qxbbw5o9vgc3hs

Answer :

1. Imported the libraries.
2. Imported test and train data from MNIST.
3. Processed the data and built the model.
4. Removed Scaling
5. Modified the model.

6. Evaluated result on test data

7. Loss and Accuracy data plot

7.Conclusions without doing scaling:

• Loss was increased from 10% to 12%
• Accuracy was reduced to 96% from 98%

---

Challenges :

Everything looks good.