Deep Learning Model Optimization and Time Series Analysis

Problem Statement

A deep learning model is performing well but the predictions are very slow. The model needs to be modified so that the performance does not deteriorate significantly but the predictions become faster without increasing the cost. Which of the given techniques should be used in this case? (Assume accuracy is not affected) Model Pruning Weight Quantization Knowledge Distillation All of the above Refer to the results of the given tests when applied to time series data. What is the stationarity of the time series data? Test: Augmented Dickey Fuller (ADF) | Result: Stationary Test: Kwiatkowski-Phillips-Schmidt-Shin (KPSS) | Result: Not stationary Trend Stationary Difference Stationary Strictly Stationary Non-stationary If the data in a regression problem contains outliers, which of the given cost functions can be used to handle outliers effectively without removing them from the dataset? Mean Squared Error (MSE) Huber Loss Log-Cosh Loss Both Huber Loss and Log-Cosh Loss Which of the followin