AI Lottery Predictions for Illinois Pick 3 Midday

Epochs

Definition: An epoch is one complete pass through the entire dataset.

Typical Range: **10 to 1000**.

• Increasing epochs can help the model learn more patterns but may cause overfitting.
• Decreasing epochs speeds training but may underfit the data.

Batch Size

Definition: How many samples are processed at once before updating the model.

Typical Range: **8 to 256**.

• Larger batch sizes speed up training but may miss finer details.
• Smaller batches are more precise but slower.

Dropout Rate

Definition: The fraction of neurons randomly ignored during training to prevent overfitting.

Typical Range: **0.1 to 0.5**.

• Higher dropout reduces overfitting but can slow learning.
• Lower dropout allows more learning but may overfit.

Learning Rate

Definition: Controls how quickly the model adjusts its knowledge each step.

Typical Range: **0.0001 to 0.1**.

• Higher rates learn faster but can skip important details.
• Lower rates learn more carefully but slowly.

Activation Function

Definition: The function that decides how a neuron fires. Introduces complexity to model patterns.

Common Choices: ReLU, Sigmoid, Tanh.

Hidden Layers

Definition: Layers between the input and output. More layers can detect more complex patterns.

Typical Range: 1 to 5.

What is Overfitting?

Overfitting happens when the model memorizes the training data rather than learning general patterns. Using dropout and proper tuning helps prevent this.