Weather Prediction on Different LSTM Techniques for Time Series Data in Chennai, India

The prediction of weather has always been an integral factor in our lives as it determines if the day of activity leads to favorable or unfavorable conditions. The existence of different deep learning models has facilitated several major problems which without such techniques would have been a very time-consuming and exhaustive process. In this following paper, we explore various types of LSTM models and analyze which among them provides the best fit for the Time Series data that is present in the Weather data. Upon gathering the datasets, a strategy called Long Short-Term Memory (LSTM) is carried out. LSTM is a deep learning model which is primarily focused on time-series data. By employing such technical improvisations on the same data with multiple models each with their own unique characteristics, we can ultimately determine the best model which has an overall better accuracy. Based on the metrics produced by employing each technique, it is found that our LSTM model which proposed here has the best performance in terms of predicting the data with an overall low error rate and loss during the training phase.