MNIST (Modified National Institute of Standards and Technology database) is the largest handwritten digits database which consists of 70,000 small images of digits. Scikit-Learn, Pytorch, and Tensorflow libraries provide helper functions to download MNIST datasets.
This dataset was collected from different high school students and employees of the US Census Bureau.
The dataset helps those who try to learn and model pattern recognition for real-world problems like:
- Recognize handwritten digits and convert those data into tables
- Convert handwritten Purchase orders values to Sales Order
- Convert handwritten stock values to post stock data
- Convert handwritten pick count to post stock transfer
- and many more…
The MNIST database holds a training set of 60,000 image examples and a testing set of 10,000 image examples. Each image example 28×28 greyscale.
This dataset helps anyone to spend less effort to create a model for your project.
Let’s write a simple code to print a handwritten digit in Scikit-Learn.
<function matplotlib.pyplot.show>
Further Reading
Posts on SAP:
Leave A Comment