The project compares multiple machine learning algorithms on digit recoginzer task. The digit recoginzer is a Kaggle competition. MNIST handwritten digit data set of computer vision. MNIST refers to Modified National institute of Standards and Technology and released in 1999. This data set is widely used for training and testing in the field of machine learning. The purpose is to identify the handwritten digit images into different digit categories from 0 to 9. In this project, I use the subset of MNIST handwritten digit data set from Kaggle Digit Recognizer competition.
The data is also from the Kaggle dataset.The Kaggle provides three csv files, train, test, and sample_submission. The train.csv is the only data file I use to train and evaluate models.
Naive bayes Classifier
Decision Tree Classifier
Support Vector Machine,
K- Nearest Neighbors
Convolutional Neural Network.
CNN and KNN perform best.
For the complete report, please read Final Report.