What hardware it will run this app on local ?
The model was implemented using the Google Colab environment with 12 GB RAM,
78 GB storage, Nvidia T4 GPU accelerator manufactured by NVIDIA Corporation, based in
Santa Clara, CA, USA and Python 3 environment. The libraries used for the implementa-
tion were the following: TensorFlow 2.17.1 and Keras 3.5.0 for building and training the
learning model; Keras Tuner 1.4.7 for hyperparameter tuning of the Keras models using
the Hyperband method; Deap 1.4 for hyperparameter tuning of the Keras models using a
genetic algorithm; and a custom implementation of the Simulated Annealing with Tensor-
Flow and Keras callbacks. The training management included using the ModelCheckpoint
callback in Keras 3.5.0 to save the best-performing model based on the validation loss, and
the ReduceLROnPlateau callback was used to adjust the learning rate when performance
stalled (factor = 0.66, patience = 5).
To identify the optimal network hyperparameter configuration, we applied a genetic
algorithm . The algorithm explored different combinations of hyperparameters, in-
cluding the number of hidden layers, number of neurons per layer, and learning rate, and
identified the optimal one.