The main focus of this study is to investigate a distributed neural network used to forecast production in Andaman and Nicobar Islands using weather parameters. The data relating to coconut yield from Central Agricultural Research Institute (CARI) have been collected for the period 1980 to 2006. The data such as average yearly rainfall, average mean temperature, relative humidity, wind speed, evaporation and sunshine hours of relevant period (1980 to 2006) have also been obtained. A multilayer perceptron with backpropagation and optimal discriminant plane method algorithm has been used. The network is trained using 17 patterns each of nine inputs. In this study, to convert nine inputs into two inputs, the nine dimensional vectors are mapped into a two-dimensional space by using a transformation. The transformed two-dimensional vector does not represent any individual feature, instead, it is a combination of nine features with no dimensional quantity.

The coconut palm is referred to as Kalpakavrishsha—The `Tree of Heaven', as each and every part of the palm is useful in one way or the other. It provides food, drink, shelter and materials for industries. Coconut is cultivated in India since ages, and it plays an important role in the social, economic and cultural activities of people.

The effort to increase the production and productivity of coconut in the country is unprecedented. This has been made possible due to the efforts of coconut cultivators, development personnel and researchers. The Andaman and Nicobar Islands consist of about 572 islands from North to South, covering an area of 8,249 km, spreading over a length of 700 km and breadth of 250 km. These islands are situated between 6° and 14° N latitude and 92° and 94° E longitude in the Bay of Bengal. The general terrain, land formation and topography of Andaman group of Islands are hilly and undulating, enclosing narrow valleys. The hot humid tropical climate prevalent in the close proximity of equator has enriched the landmass with evergreen forests. The average annual rainfall in these islands is 3,000 mm, mean relative humidity ranges from 80% to 90% and mean minimum temperature is 23.2 °C and mean maximum temperature is 30.7 °C.

Agriculture in these islands is barely 150 years old and the agricultural scenario is not very much encouraging. The green revolution could not usher in these islands due to various constraints. The climatic condition of these islands is favorable for cultivation of all types of tropical crops, but it is also to be noted that pests and diseases often occur. Hence, a single crop is not fruitful. Therefore, the concept of multi-tier/multiple cropping system has come into practice for sustainable agricultural production to the farmers of these islands.

An Artificial Neural Network (ANN) is an information processing model based upon the functioning of neurons, or nerve cells, found in the human brain and nervous system. Researchers are able to create an information processing system for forecasting that operates in the manner of the human nervous system. Scientists have learned that the human brain holds over 100 billion neurons connected by fibers that form networks which are used to transfer signals and process information.

Computer Sciences Journal, Coconut Yield Prediction, Neural Networks, Optimal Discriminant Plane Method Algorithm, Artificial Neural Network, Human Nervous System, Green Revolution, Agricultural Production, Information Processing System, Central Agricultural Research Institute, CARI, Linear Mapping Algorithms.