ORIGINAL ARTICLE STUDIES ON LEARNING IN SIMPLE NEURONS AND MULTILAYER PERCEPTRON USING HYDROLOGICAL FACTOR

R Manimegalai, V Jayaraj
2016 Int. J. Modn. Res. Revs   unpublished
1.INTRODUCTION The multilayer perceptron is the most known and most frequently used type of neural network. On most occasions, the signals are transmitted within the network in one direction: from input to output. There is no loop, the output of each neuron does not affect the neuron itself. There are also feedback networks, which can transmit impulses in both directions, due to reaction connections in the network. These types of networks are very powerful and can be extremely complicated. They
more » ... y complicated. They are dynamic, changing their condition all the time, until the network reaches an equilibrium state, and the search for a new balance occurs with each input change. Evaporation and transpiration is an essential component of the hydrological cycle, and its accurate estimation forms the basis for irrigation requirements. A new approach to determine-he reference crop Evaporation and transpiration has been proposed and developed, that employs the pattern matching capability of ANN. Evaporation and transpiration is one of the important components of the hydrological cycle. An accurate estimate of it is essential for the hydrological water-balance, irrigation, and water resources planning and management (Raghuwanshi, et al) with the emphasis of management practices for optimal use of water,crop water requirement forms a vital role in the planning,design and operation of water resource systems. Potential Evapotranspiration (PET) is defined by penman as Evaporation and transpiration from an extended surface of short green crop, actively growing, completely shading the ground of uniform height and not short of water. Mohan used the term REF_ET instead of PET for potential Evapo transpiration of a reference crop A large number of methods have been developed and tested for estimation of RET_ET for varying geographic and tested for estimation of RET_ET for a varying geographic and climatologic conditions from simple empirical equation to complex methods. Rahuvanshi and Wallender, (1998) have recommended penman method for reliable estimation of REF_ET in India. Amatya et.el reported that penman method is quiet reliable when the necessary weather and vegetation data are known, but these I/P ' s are very expensive to obtain. In the present study, an attempt has been made to employ the pattern matching ability of neural networks to simplify the estimation of RET_ET. Given the preliminary nature of the research work being reported here, there will be no attempt to recommend the replacement of existing models for estimation of RET_ET based on the hypothesis of improved easiness and accuracy of the ANN models. The main purpose of the present study is to demonstrate promise and feasibility of ANN networks being applied to REF_ET estimation. ABSTRACT The objective of this study is to observe the simple learning in neuron and multilayer perception. The potential of an artificial neural network to perform simple learning in neuron and multilayer is examined. The multilayer perceptron is often quite slow, requiring thousands or tens of thousands of epochs for complex problems and simple learning in neuron is supervised learning, unsupervised learning and reinforced learning. The present study represents the possibility to observe learning in simple neurons and multilayer perceptron by using hydrological factor.
fatcat:qxkobmhxbbbgbab25drjjzb5uq