Build a machine learning model for stock price change point prediction using MODWT
Budget: ₹1,500 – ₹12,500 INR
1. The original series is broken into a number of distinct sub-series using MODDWT (sometimes referred to as wavelets).
2. Each sub-series in the hybrid MODWT-RNN models is projected separately using RNN.
3. The predicted sub-series are once again blended to obtain the final forecast value.
4. To calculate performance/error measures, the final prediction value (obtained in the previous phase) is compared to the original series.
5. Using MAPE, MAE, RMSE, and MDA to evaluate the effectiveness of the models provided, Arima and RNN-LSTM are utilized to compare MODWT-ability RNN's for prediction.
2. Each sub-series in the hybrid MODWT-RNN models is projected separately using RNN.
3. The predicted sub-series are once again blended to obtain the final forecast value.
4. To calculate performance/error measures, the final prediction value (obtained in the previous phase) is compared to the original series.
5. Using MAPE, MAE, RMSE, and MDA to evaluate the effectiveness of the models provided, Arima and RNN-LSTM are utilized to compare MODWT-ability RNN's for prediction.