SECRET WRITING BY ARTIFICIAL NEURAL NETWORK Neetu Singh 1 and Kanika Khetarpal 2 I. INTRODUCTION Artificial Neural Network system is developed to process information similar to the human nervous system. Its architecture is modeled according to the architecture of actual human neural network system. The key part of its architecture is the design of how it processes the information. A neural network consists of a large number of neurons with complex interconnectivity. This complex network works with unity to solve a problem. Similar to human neural network, artificial neural networks also learns using examples. Different training sets or examples can train ANN for different applications like pattern recognition, cryptography, data classification etc. In a biological neural network, the synapses that connect neurons adjust based on the learning experience. ANN also functions similarly. There are numerous advantages of an artificial neural network but most significant is that it learns by observing the information. This helps in using this network as a random approximation tool. This leads to developing a system to reach the solution in a most ideal and cost-effective way. Artificial neural networks are efficient in saving time and money by processing the information and reaching to an optimum solution using training examples instead of whole data sets. There are three layers in an artificial neural network. First is input layer to gather the data for information processing. Second is a hidden layer to process the information and reach to an output. The third layer which provides with the output. The description is depicted in figure 1. Fig.1 ANN Structure 1 Department of Computer Science Engineering Guru Tegh Bahadur Institute Of Technology, New Delhi, India 2 Department of Computer Science Engineering Guru Tegh Bahadur Institute Of Technology, New Delhi, India International Journal of Latest Trends in Engineering and Technology Vol.(8)Issue(2), pp.314-320 DOI: http:/ / dx.doi.org/ 10.21172/ 1.82.042 e-ISSN:2278-621X Abstract—Cryptography is the practice or study of transmitting sensitive information in an encoded format over a network channel. In the age of modern technology cryptography has a key role to play. Huge amounts of information gets transmit every day over numerous networks that contain all kinds of personal and private data. This data is needed to be safeguarded from getting exploited by unauthorized people therefore a strong cryptographic system is needed. The paper focuses on the ways to achieve that goal by using the technique of artificial neural network in cryptography and to explore how this technique can be utilized as a powerful tool to protect the information from unauthorized users. Key words: Artificial Neural Network (ANN), Neurons, Sigmoid Function, Dendrite, Axon, Synapse.