Example 1 of D4.3 - Deep Learning

[cozzolinoac11]

Use case

common

Name of resource

LeNet Classifier

ID

lenet_classifier

Description

Multi-layer Convolutional Neural Network for image classification

Main category

Deep Learning

Other category

No response

Publication date

2023-04-04

Objective

classification

Platform

Google Colab

Framework

Keras

Architecture

CNN - Convolutional-Neural-Network

Approach

supervised

Algorithm

LeNet

Processor

gpu

OS

linux

Keyword

classification, CNN, LeNet

Reference link

https://en.wikipedia.org/wiki/LeNet

Example

https://github.com/cozzolinoac11/wildfire_prediction/blob/main/ann.ipynb

Input data used

1. https://public.epsilon-italia.it/FAIRiCUBE/wildfire-classification/data_numpy.zip

Characteristics of input data

1. Numpy arrays. (Perfectly) balanced classes.

Biases and ethical aspects

No response

Output data obtained

1. http://www.epsilon-italia.it/public/model.zip

Characteristics of output data

1. Keras model for wildfire or nowildfire classification. The model gets in input a dataset as numpy arrays (dimension 100x100x3) and returns the predicted labels.

Performance

Accuracy score: 0.9505 (validation). Running time: 2 min for 23 training epochs with early stopping (total number of epochs: 50) on a gpu Nvidia a100. Modified hyperparameters: Input shape: (100,100,3); Optimizer: 'adam'; batch size: 128. Train-test-valid split: 70-15-15. Loss function: sparse_categorical_crossentropy.

Conditions for access and use

cc-by-4.0

Constraints

No response