Abstract: Various arrangements are presented for training and using a machine learning model. A first training data set may be created that has more samples but fewer dimensions than a second dataset. A second set of training data, created from the second dataset, has at least one additional dimension of data than the first set of training data. An additional dimension of data can then be simulated for the first set of training data. The simulated additional dimension of data can be incorporated with the first set of training data. A first machine learning model can be trained based on the first set of training data that comprises the simulated additional dimension of data to obtain various weights. A second machine learning model can then be trained based on the second set of training data and the obtained plurality of weights from the first trained machine learning model.

Abstract: A multi-tone transceiver with a components forming a transmit path and a receive path configured to couple via a subscriber line to an opposing multi-tone transceiver for frequency division multiplexed multi-tone modulated communications therewith is disclosed. A noise margin channel identifier is configured to identify within a received tone set, discrete tones each associated with a corresponding one of at least two channels differing from one another in a relative noise margin of associated tones. A Viterbi decoder is responsive to the channel identification provided by the noise margin channel identifier to discretely decode each of the at least two channels; thereby improving the fidelity of the error correction provided by the Viterbi decoder by discretely processing the identified channels within the received set of tones.