Supervised Machine Learning Model for facial recognition.
TASKS PERFORMED: Task 1: #Loading data The first task is to load the images from data/folder as NumPy array. First manually go through the data, and understand how it is organised.
X should be a NumPy array containing images and y should contain the corresponding labels.
Task 2: #Analysing data Visualise the first image of our dataset.
Task 3: #Feature Extraction Currently, X is a 3D NumPy array. For each of the 400 images, we have a corresponding 2D matrix storing the pixel values. We will flatten this 2D matrix and use the flattened vector as the feature vector for the image.
Task 4: #Dimensionality Reduction #Data Visualisation. Reshape X into 400xN, where N is the number of features (112x92). Task 5: #Speeding up training and reducing memory #Splitting the dataset #Standardising the Dataset #Training and Hyperparameter Tuning Task 6: #Make an object of PCA class named pca so that 99% of variance is preserved. #Testing Task 7: #Predictions