Wide & Deep Learning Model for CTR Prediction

This application trains a Wide & Deep model using a fixed filtered_train.csv dataset. Configure model parameters and observe training loss, accuracy, and predictions.

Number of Epochs

1 200

Batch Size

16 256

Deep Network Hidden Layers (Neurons)

Layer 1 Neurons

Layer 2 Neurons (set to 0 to disable)

Layer 3 Neurons (set to 0 to disable)

Dropout Rates (0.0 to 1.0)

Dropout Layer 1

0 0.9

Dropout Layer 2

0 0.9

Dropout Layer 3

0 0.9

Status and Real-time Loss

Test Accuracy

Prediction Comparison (First 10 Samples)