Image Classification with TensorFlow

Task: Built a neural network model to solve a basic image classification problem, on MNIST dataset.

Dataset: MNIST Dataset

Architecture: Custom

Implementation Steps:

1. Import Packages (numpy, tensorflow, matplotlib)

2. Load Data (load the data and split the data into train and test)

3. One Hot Encoding (labels)

4. Data Reshaping

5. Data Normalization (Mean and Standard Deviation)

6. Creating a model (structure, layers)

7. Compiling the model (optimizer, loss, metrics)

8. Model fitting

9. Evaluating the model (loss , accuracy)

10. Predictions

11. Plotting (Matplotlib)

Details: Loss - Categorical Crossentropy, Optimizer - SGD

Results: Accuracy 97%

GitHub Code: https://github.com/Rubo12345/Image-classification-with-TensorFlow