Competition: MNIST

Programming Language: Python

Accuracy: 99.21%

Place: 462 / 1621

Kaggle is a great platform for AI developers and data scientists to collaborate and meet each other. It has different features, like searching for datasets, participating in competitions, online interpreters and more. In my time, I have enrolled in Kaggle competitions and used online datasets for practicing my skills.

Participating in Kaggle Competitions is a developing and interesting experience. Through them you can practice and gather new knowledge, refining your coding and data analysis skills. Kaggle competitions also allow you to form new connections and meet other people, sharing your knowledge and getting better all together.

In recent times, I have competed in the Kaggle MNIST competition. MNIST is a dataset consisting of 60000 images of hand-written numbers and their labels, so the main point of the competition would be to create and train a model that can correctly name what number is written on the image. With my knowledge, I have managed to get a 99.21% accuracy on the test dataset, ending up in 462 place out of 1600 people. With this competition, I have learned what architecture should I choose and refined my analytical skills.