A computational study reveals that lightweight machine learning models paired with trend-based fundamental feature ...
A framework combining discrete wavelet transforms and temporal convolutional networks predicted continuous stock trends with 72.85 percent accuracy across 400 equities.