Description
WRITE PYTHON CODE**Objective:** The objective of this assignment is to practice data preprocessing techniques for preparing an image dataset for training a Convolutional Neural Network (CNN). You will go through the steps of data collection, splitting, resizing, normalization, and optional data augmentation. **Instructions:** **Data Collection and Organization (5 points):** – Find an image dataset of your choice. It could be a dataset for image classification, object detection, or any other image-based task. – Download and organize the dataset, ensuring it’s correctly labeled and categorized. **Data Splitting (5 points):** – Split the dataset into three subsets: training, validation, and test sets. Use a reasonable split ratio (e.g., 70% for training, 15% for validation, and 15% for testing). **Image Resizing (5 points):** – Choose a CNN architecture that you plan to use for your task. Check the input size expected by the chosen model. – Write a Python script to resize all images in your dataset to match the input size of the chosen CNN model. **Normalization (10 points):** – Implement data normalization. Write Python code to normalize the pixel values of your images. – Ensure that the pixel values are within the range [0, 1] by either dividing by 255 (if the original values are 0-255) or by subtracting the mean and dividing by the standard deviation. **Data Augmentation (Optional, 5 points):** – Implement data augmentation techniques to increase the diversity of your training data. You can include operations such as random rotation, random cropping, and horizontal flipping. – Write Python code to apply these data augmentation techniques to the training images. **Data Batching (5 points):** – Organize your dataset into batches. Define batch sizes suitable for your available computing resources and model. **One-Hot Encoding (for Classification, 5 points):** – If your task is image classification, write Python code to one-hot encode the class labels of your dataset. **Test Your Preprocessed Data (5 points):** – Write Python code to visualize a few images from your preprocessed training data. – Ensure that the images are correctly resized, normalized, and optionally augmented.
![2a. Write a Python program to perform supervised classification on the Iris and Indian Pines datasets using Naive Bayes, Support Vector Machines (with RBF kernel), and Multi-layer Perceptron (MLP) classifiers for training sizes ={10%,20%,30%,40%,50%} for each of the below cases: i) With dimensionality reduction - Reduce data based on your choice of 'K' dimensions from 1a using each of the dimensionality reduction methods (PCA) followed by supervised classification by the listed classifiers. ii) Without dimensionality reduction - Data is followed by supervised classification using the listed classifiers. iii) Provide the plots for overall training accuracy, and overall classification accuracy versus the training size for all methods (classification schemes). Tabulate the classwise classification accuracies (i.e. extension of the sensitivity and specificity values) only for 30% training size over all methods for each dataset for case ii) i.e. without dimensionality reduction. Please note that the point distribution is as follows: - Total 4 plots, i.e. 2 plots for each dataset [case i and ii, overall training, and testing accuracy] - 2 tables (classwise accuracy), i.e, 1 table per dataset - Total points available: 30 points (5 points per plot/table)](https://gotit-pro.com/wp-content/uploads/2023/10/638be90e-4e3d-41db-a9d6-5279466b2f80.jpg)
Alice Elkins –
Tutor is patient, highly efficient and well spoken. I can always count on early delivery! Thx!!
Narayana Rao –
Fantastic support for every assignment, every time. 5+ Stars!
Carmen Hoggard –
Thanks a lots, I really appreciate your effort it is an excellent report.
Michael Jones –
Great job, very quick and informative! Done in an hours! Thank you for a job well done.
Chris Brooks –
Thank for your help and service. I really like the work that they did on my paper.