Scribd
Upload
12 views1 page

Model Training & Hyperparameter Tuning

The document outlines a practical session focused on model training and hyperparameter tuning, emphasizing the importance of selecting optimal parameters for various classifiers. It details the training process for different models, including Naive Bayes, image classifiers, and regression models, and highlights the best parameters found during tuning, such as using Ridge regularization with specific values. The session concludes with evaluation metrics, noting an initial accuracy of 97% and improvements achieved through tuning.

Uploaded by

RAYUDU SRD
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
12 views1 page

Model Training & Hyperparameter Tuning

The document outlines a practical session focused on model training and hyperparameter tuning, emphasizing the importance of selecting optimal parameters for various classifiers. It details the training process for different models, including Naive Bayes, image classifiers, and regression models, and highlights the best parameters found during tuning, such as using Ridge regularization with specific values. The session concludes with evaluation metrics, noting an initial accuracy of 97% and improvements achieved through tuning.

Uploaded by

RAYUDU SRD
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Practical Session: Model Training and Hyperparameter Tuning

1. Introduction

 Understand the process of training a model and the importance of hyperparameter tuning.

 Explore different classifiers and their performance based on various parameters.

2. Models and Classifiers

 Naive Bayes Classifier: A probabilistic classifier based on applying Bayes' theorem.

 Image Classifier: A model used to classify images into different categories.

 Regression Model: A model predicting a continuous output variable.

3. New Model Training

 Train the new model on a given dataset.

 Evaluate the model's performance.

4. Parameter Tuning

 Identify and apply the best hyperparameters for the model.

5. Extracting Parameters

 Identify the best parameters from multiple attempts.

6. Key Parameters and Results

 Gaussian Parameters: Use Gaussian distribution parameters.

 Maximum Iterations: Set to 100.

 Penalty (Regularisation): Identify the best regularization technique.

o Ridge Regularisation: Found to perform well when C=1C = 1 and γ=0\gamma = 0.

 Training Results:

o Best results achieved with the following parameters:

 Maximum Iterations: 100

 Regularisation: Ridge

 Parameter γ\gamma: 0

7. Evaluation Metrics

 Accuracy: Measure the accuracy after hyperparameter tuning.

o Initial Accuracy: 97%

o Improved Accuracy: Achieved after hyperparameter tuning and training with


different parameters.

You might also like