Browsing by Author "Nanjundiah, R.S."

Now showing 1 - 3 of 3

Deep Learning Based LSTM and SeqToSeq Models to Detect Monsoon Spells of India
(2019) Viswanath, S.; Saha, M.; Mitra, P.; Nanjundiah, R.S.
Monsoon spells are important climatic phenomenon modulating the quality and quantity of monsoon over a year. India being an agricultural country, identification of monsoon spells is extremely important to plan agricultural policies following the phases of monsoon to attain maximum productivity. Monsoon spells� detection involve analyzing and predicting monsoon at daily levels which make it more challenging as daily-variability is higher as compared to monsoon over a month or an year. In this article, deep-learning based long short-term memory and sequence-to-sequence models are utilized to classify monsoon days, which are finally assembled to detect the spells. Dry and wet days are classified with precision of 0.95 and 0.87, respectively. Break spells are observed to be forecast with higher accuracy than the active spells. Additionally, sequence-to-sequence model is noted to perform superior to that of long-short term memory model. The proposed models also outperform traditional classification models for monsoon spell detection. � Springer Nature Switzerland AG 2019.
Deep Learning Based LSTM and SeqToSeq Models to Detect Monsoon Spells of India
(Springer Verlag service@springer.de, 2019) Saicharan, S.; Saha, M.; Mitra, P.; Nanjundiah, R.S.
Monsoon spells are important climatic phenomenon modulating the quality and quantity of monsoon over a year. India being an agricultural country, identification of monsoon spells is extremely important to plan agricultural policies following the phases of monsoon to attain maximum productivity. Monsoon spellsâ€™ detection involve analyzing and predicting monsoon at daily levels which make it more challenging as daily-variability is higher as compared to monsoon over a month or an year. In this article, deep-learning based long short-term memory and sequence-to-sequence models are utilized to classify monsoon days, which are finally assembled to detect the spells. Dry and wet days are classified with precision of 0.95 and 0.87, respectively. Break spells are observed to be forecast with higher accuracy than the active spells. Additionally, sequence-to-sequence model is noted to perform superior to that of long-short term memory model. The proposed models also outperform traditional classification models for monsoon spell detection. Â© Springer Nature Switzerland AG 2019.
The slow-manifold for the Lorenz-Krishnamurthy model has been studied. By minimizing the evolution rate we find that the analytical functions for the fast variables are devoid of high frequency oscillations. However upon solving this model with initial values of the fast variables obtained from the analytical functions, the LK model exhibits high frequency oscillations. Upon using the time derivatives of the analytic functions for computing the evolution of fast variables, we find a slow-manifold in the neighbourhood of the LK model. Minimization of evolution rate does not guarantee the invariance of the manifold. Using a locally linear approximate reduction scheme, the invariance can be maintained. However, the solutions so obtained do develop high frequency oscillations. The onset of these high frequency oscillations is delayed vis-a-vis other previous studies. These methods have potential to be used in improving the predictions of weather systems.
(Southwest Missouri State University, Revisiting the slow manifold of the Lorenz-Krishnamurthy quintet) Motamarri, M.; Nanjundiah, R.S.; Vasudeva Murthy, A.S.
2006