• Rani M T
• Dr Siddaraju

The MNIST database provides huge number of opportunities to learn and work using many techniques of Machine Learning. One of those an Artificial Neural Network place an important role in analyzing the database. The MNIST database has 70000 data points, of which 60000 is train data set and 10000 is test data set. There are some work has been carried using hardware components like FPGA, and CNN which is hardware intensive and more matrix complexity. This paper explains the combination of ReLU-Sigmoid and Leaky ReLU-Sigmoid with the simulation results in achieving the performance above 96% on a software platform (Jupyter Notebook).

ANN, ReLU, Leaky-ReLU, Sigmoid

"Performance Analysis of Artificial Neural Network on Digit Recognition using MNIST Database", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 1, page no.f77-f83, January-2022, Available :http://www.jetir.org/papers/JETIR2201514.pdf

"Performance Analysis of Artificial Neural Network on Digit Recognition using MNIST Database", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 1, page no. ppf77-f83, January-2022, Available at : http://www.jetir.org/papers/JETIR2201514.pdf