vinicioslugli/2023-2b-t2-m8-p6

2023 - Módulo 8 - Ponderada 6

2023-2B-T2-M8-P6

Quick Starter

This quick starter provides a simple example of using the MNIST dataset with TensorFlow and Keras. It includes a Convolutional Neural Network (CNN) for digit classification, training the model, and evaluating its performance.

Requirements

Ensure you have the required dependencies installed by running:

Currently using Python 3.11

pip install -r requirements.txt

Usage

1. Run the main.py file to train the model and execute predictions on the test dataset.

python main.py

Code Overview

The code in main.py contains a class MNISTClassifier encapsulating the MNIST model and related functions. It includes the following methods:

• train_model(epochs): Trains the model on the MNIST training dataset for the specified number of epochs.
• evaluate_model(): Evaluates the trained model on the MNIST test dataset and prints the test accuracy.
• predict_samples(num_samples): Displays predictions for a specified number of samples from the test dataset.

Additionally, the code includes a test class TestMNISTClassifier within which there is a test method test_model_evaluation. This method checks if the model is initialized, trains it, evaluates its accuracy, and predicts samples, ensuring the accuracy is above a specified threshold.

MNISTClassifier Class

The MNISTClassifier class utilizes TensorFlow and Keras to define, compile, and train a Convolutional Neural Network for digit classification on the MNIST dataset. The architecture consists of a convolutional layer, max-pooling layer, flattening layer, and dense layers.

Notes

• The test accuracy is checked to ensure it is greater than 95%.
• Predictions for sample images are displayed using Matplotlib.

Demo

demo.webm