Transfer learning

In 1993, Lorien Pratt published a paper on transfer in machine learning, formulating the discriminability-based transfer (DBT) algorithm.[5]

In 1997, the journal Machine Learning published a special issue devoted to transfer learning,[6] and by 1998, the field had advanced to include multi-task learning,[7] along with a more formal analysis of its theoretical foundations.[8] Learning to Learn,[9] edited by Pratt and Sebastian Thrun, is a 1998 review of the subject.

Transfer learning has also been applied in cognitive science, with the journal Connection Science publishing a special issue on reuse of neural networks through transfer in 1996.[10]

The definition of transfer learning is given in terms of domain and task. The domain ${\displaystyle {\mathcal {D}}}$ consists of: a feature space ${\displaystyle {\mathcal {X}}}$ and a marginal probability distribution ${\displaystyle P(X)}$, where ${\displaystyle X=\{x_{1},...,x_{n}\}\in {\mathcal {X}}}$. Given a specific domain, ${\displaystyle {\mathcal {D}}=\{{\mathcal {X}},P(X)\}}$, a task consists of two components: a label space ${\displaystyle {\mathcal {Y}}}$ and an objective predictive function ${\displaystyle f(\cdot )}$(denoted by ${\displaystyle {\mathcal {T}}=\{{\mathcal {Y}},f(\cdot )\}}$), which is learned from the training data consisting of pairs, which consist of pairs ${\displaystyle \{x_{i},y_{i}\}}$, where ${\displaystyle x_{i}\in X}$ and ${\displaystyle y_{i}\in {\mathcal {Y}}}$. The function ${\displaystyle f(\cdot )}$ can be used to predict the corresponding label,${\displaystyle f(x)}$, of a new instance ${\displaystyle x}$.[11]

Given a source domain ${\displaystyle {\mathcal {D}}_{S}}$ and learning task ${\displaystyle {\mathcal {T}}_{S}}$, a target domain ${\displaystyle {\mathcal {D}}_{T}}$and learning task ${\displaystyle {\mathcal {T}}_{T}}$, transfer learning aims to help improve the learning of the target predictive function ${\displaystyle f_{T}(\cdot )}$ in ${\displaystyle {\mathcal {D}}_{T}}$ using the knowledge in ${\displaystyle {\mathcal {D}}_{S}}$ and ${\displaystyle {\mathcal {T}}_{S}}$, where ${\displaystyle {\mathcal {D}}_{S}\neq {\mathcal {D}}_{T}}$, or ${\displaystyle {\mathcal {T}}_{S}\neq {\mathcal {T}}_{T}}$.[11]

Algorithms are available for transfer learning in Markov logic networks[12] and Bayesian networks.[13] Transfer learning has also been applied to cancer subtype discovery,[14] building utilization,[15][16] general game playing,[17] text classification,[18][19] digit recognition[20] and spam filtering.[21]

In 2020 it was discovered that, due to their similar physical natures, transfer learning is possible between Electromyographic (EMG) signals from the muscles when classifying the behaviours of Electroencephalographic (EEG) brainwaves from the gesture recognition domain to the mental state recognition domain. It was also noted that this relationship worked vice-versa, showing that EEG can likewise be used to classify EMG in addition[22]. The experiments noted that the accuracy of neural networks and convolutional neural networks were improved through transfer learning both at the first epoch (prior to any learning, ie. compared to standard random weight distribution) and at the asymptote (the end of the learning process). That is, algorithms are improved by exposure to another domain.

• Crossover (genetic algorithm)
• Domain adaptation
• General game playing
• Multi-task learning
• Multitask optimization

• Thrun, Sebastian; Pratt, Lorien (6 December 2012). Learning to Learn. Springer Science & Business Media. ISBN 978-1-4615-5529-2.CS1 maint: ref=harv (link)