given a stockmarket time series, draw support and resistance trendlines above and below the current channel. label swing points.
recommend using python trendln repo. this will be used as training for the next part. plot results and save all data in unique filenames in a subdirectory
using the above train a network, likely a boosted essemble to predict given a time series the levels of the support resistance points above and below a point on the time series. identify suitable indicators to be added for training, sma, hurst etc . produce appropriate metrics from training
ability to save and restore model and run predictions and run hyper parameter search
data will be provided. prefer pytorch