predict_window_momentum_LOAD: predicts momentum for asset according to LOAD strategy
In benSepanski/pfselect:

predicts next momentum for one asset given previous prices according to LOAD strategy (see paper mentioned in description of backtest_LOAD), momentu is defined in evaluate_momentum_at_window.

predict_window_momentum_LOAD(
  prev_prices,
  regularization_factor,
  momentum_threshold,
  min_var = .Machine$double.eps^0.5
)

`prev_prices`	a numeric vector of previous prices
`regularization_factor`	λ in the referenced paper, the regularization coeffecient for weight decay when regressing the prices in the recent time window. Must be non-negative
`momentum_threshold`	η in the referenced paper. If the regressed slope is greater than η we say the stock has momentum. This must be greater than 0.
`min_var`	(OPTIONAL) minimum variance to be determined non-constant