1. Create a train-test split and classify the images using any classifier you have used previously. What is the classifier performance?

You want to classify images of dogs from cats. You have collected 2,000 images of dogs and 2,000 images of cats. How would you split the data effectively into a training set and a validation set?Group of answer choicesYou should split by light or dark fur color.You should split by whether the image contains a cat or dog.You should split by high- or low-quality images.You should split uniformly at random.

The dataThe dataset we will be using is called MNIST. This is a large collection of hand-drawn digits 0 to 9 and is a good dataset to learn image classification on as it requires little to no preprocessing.The dataset can be downloaded from The MNIST Database. Download all four files. These files are the images and their respective labels (normally, we're required to split the x (image data / characteristics) and y (labels) out during preprocessing, but this has already been done for us). The dataset has also already been split into a train and a test set.Once you've downloaded the data, make sure that the data are in the same folder as this Jupyter notebook. If you've managed to do all that, we can now begin!By default, the MNIST files are compressed in the gzip format. The following two functions will extract the data for you. ** Don't change this code. **In [2]:def extract_data(filename, num_images, IMAGE_WIDTH): """Extract the images into a 4D tensor [image index, y, x, channels].""" with gzip.open(filename) as bytestream: bytestream.read(16) buf = bytestream.read(IMAGE_WIDTH * IMAGE_WIDTH * num_images) data = np.frombuffer(buf, dtype=np.uint8).astype(np.float32) data = data.reshape(num_images, IMAGE_WIDTH*IMAGE_WIDTH) return data​def extract_labels(filename, num_images): """Extract the labels into a vector of int64 label IDs.""" with gzip.open(filename) as bytestream: bytestream.read(8) buf = bytestream.read(1 * num_images) labels = np.frombuffer(buf, dtype=np.uint8).astype(np.int64) return labelsChallenge 1: Extracting the dataThe MNIST dataset consists of 60,000 training images and 10,000 testing images. This is a lot of data! Let's not extract all of that right now. Create a function get_data that uses the above functions to extract a certain number of images and their labels from the gzip files.The function will take as input two integer values, the number of train and test images to be extracted. Let's extract 5000 train images and 1000 test images. The function then returns four variables in the form of (X_train, y_train), (X_test, y_test), where (X_train, y_train) are the extracted images and labels of the training set, and (X-test, y_test) are the extracted images and labels of the testing set. (Hint – you'll have to use the functions provided more than once.)Image pixel values range from 0 to 255. We need to normalise the image pixels so that they are in the range 0 to 1.Function specifications:Should take two integers as input, one representing the number of training images and the other the number of testing images.Should return two tuples of the form (X_train, y_train), (X_test, y_test).Note that the size of the MNIST images is 28x28.Usually when setting up your dataset, it is a good idea to randomly shuffle your data in case your data are ordered. Think of this as shuffling a pack of cards. Here, however, we aren't going to shuffle the data so that all our answers are the same.

You are fine-tuning a support vector machine (SVM) classifier to categorise images based on their content. The dataset consists of various animal images, and you suspect that different kernel functions might yield better classification accuracy. You decide to test which SVM kernel—linear or radial basis function (RBF)—works best for your specific dataset. Below is your initial code setup:from sklearn.svm import SVCfrom sklearn.datasets import load_digitsfrom sklearn.model_selection import train_test_splitfrom sklearn.metrics import accuracy_score# Load a dataset of digit imagesdigits = load_digits()X = digits.datay = digits.target# Split the data into training and testing setsX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=42)# Initialise two SVM classifiers, one with a linear kernel and another with an RBF kernelsvm_linear = SVC(kernel='linear')svm_rbf = SVC(kernel='rbf')# [Your Code Here] - Train both classifiers on the training data# [Your Code Here] - Predict the test set results with both classifiers# [Your Code Here] - Calculate and print the accuracy scores for both classifiersWhich of the following options correctly completes the task of training both SVM classifiers, predicting the test set results, and calculating the accuracy for eachsvm_linear.train(X_train, y_train)svm_rbf.train(X_train, y_train)y_pred_linear = svm_linear.classify(X_test)y_pred_rbf = svm_rbf.classify(X_test)print("Linear Kernel Accuracy:", accuracy_score(y_test, y_pred_linear))print("RBF Kernel Accuracy:", accuracy_score(y_test, y_pred_rbf))svm_linear.fit(X_train, y_train)svm_rbf.fit(X_train, y_train)y_pred_linear = svm_linear.predict(X_test)y_pred_rbf = svm_rbf.predict(X_test)print("Linear Accuracy:", accuracy_score(y_test, y_pred_linear))print("RBF Accuracy:", accuracy_score(y_test, y_pred_rbf))svm_linear.fit(X_train, y_train)y_pred_linear = svm_linear.predict(X_train)svm_rbf.fit(X_train, y_train)y_pred_rbf = svm_rbf.predict(X_train)print("Accuracy with Linear Kernel:", accuracy_score(y_train, y_pred_linear))print("Accuracy with RBF Kernel:", accuracy_score(y_train, y_pred_rbf))svm_linear.fit(X_train, y_train)y_pred_linear = svm_linear.predict(X_test)svm_rbf.fit(X_train, y_train)y_pred_rbf = svm_rbf.predict(X_test)print("Accuracy with Linear Kernel:", accuracy_score(y_test, y_pred_linear))print("Accuracy with RBF Kernel:", accuracy_score(y_test, y_pred_rbf))

Question 4You are working to create an object detection system, like the ones described in the lectures, to locate cats in a room. To have more data with which to train, you search on the internet and find a large number of cat photos.Which of the following is true about the system?1 point

What is the primary advantage of using CNNs in image recognition compared to traditional computer vision techniques?a.Improved feature engineeringb.Better performance with small datasetsc.Automatic feature learningd.Faster execution on CPU